Настройки

Укажите год
-

Небесная энциклопедия

Космические корабли и станции, автоматические КА и методы их проектирования, бортовые комплексы управления, системы и средства жизнеобеспечения, особенности технологии производства ракетно-космических систем

Подробнее
-

Мониторинг СМИ

Мониторинг СМИ и социальных сетей. Сканирование интернета, новостных сайтов, специализированных контентных площадок на базе мессенджеров. Гибкие настройки фильтров и первоначальных источников.

Подробнее

Форма поиска

Поддерживает ввод нескольких поисковых фраз (по одной на строку). При поиске обеспечивает поддержку морфологии русского и английского языка
Ведите корректный номера.
Ведите корректный номера.
Ведите корректный номера.
Ведите корректный номера.
Укажите год
Укажите год
Применить Всего найдено 565. Отображено 100.
04-07-2023 дата публикации

Method and system for protecting semantic privacy of workers in spatial crowdsourcing

Номер: CN116383878A
Автор: ZHOU XU, CHEN SI, YANG ZHIBANG, REN TIANYUE, LI KENLI
Принадлежит:

The invention discloses a method for protecting semantic privacy of workers in spatial crowdsourcing, which comprises the following steps of: acquiring a regional point data set, determining corresponding basic semantics, counting and classifying point position semantics corresponding to the point data set according to the point data set and the basic semantics, constructing a three-layer semantic tree model according to counting and classifying results, and constructing the semantic privacy of the workers according to the three-layer semantic tree model. The method comprises the following steps: selecting a three-layer semantic tree model, determining a semantic weight corresponding to the three-layer semantic tree model and a grid semantic similarity, obtaining worker position information in a selected area according to the three-layer semantic tree model and the grid semantic similarity, constructing a privacy space decomposition structure according to the worker position information ...

Подробнее
Номер записи: 1
16-11-2023 дата публикации

DETECTING VEHICLE APERTURE AND/OR DOOR STATE

Номер: US20230368660A1
Автор: Mahsa Ghafarianzadeh, Benjamin John Sapp
Принадлежит:

A method and system of determining whether a stationary vehicle is a blocking vehicle to improve control of an autonomous vehicle. A perception engine may detect a stationary vehicle in an environment of the autonomous vehicle from sensor data received by the autonomous vehicle. Responsive to this detection, the perception engine may determine feature values of the environment of the vehicle from sensor data (e.g., features of the stationary vehicle, other object(s), the environment itself). The autonomous vehicle may input these feature values into a machine-learning model to determine a probability that the stationary vehicle is a blocking vehicle and use the probability to generate a trajectory to control motion of the autonomous vehicle.

Подробнее
Номер записи: 2
08-08-2023 дата публикации

Crop yield estimation method, system, equipment and medium

Номер: CN116562446A
Автор: LIEW HOW KIAT, HE LIUPENG, WANG FENGZUO, WEI WEI, DING YIBO, JIANG CHENGZHEN, HU JIE, LI NING
Принадлежит:

The invention discloses a hybrid model construction method and system for estimating crop yield, equipment and a medium, and relates to the field of precision agriculture. The method comprises the following steps: acquiring a historical yield influence factor and a historical crop yield of a region to be measured, decomposing the historical crop yield into a trend yield and a climate yield, and determining a relative climate yield according to the trend yield and the climate yield; based on an empirical statistical model, analyzing the correlation between the historical yield influence factor and the relative climate yield, determining a key growth period, and screening an optimal yield influence factor according to the key growth period; the optimal yield influence factor is a historical yield influence factor in a key growth period; according to the optimal yield influence factor, the relative climate yield and the crop yield, constructing a random forest regression model to estimate ...

Подробнее
Номер записи: 3
28-07-2023 дата публикации

Ternary lithium battery capacity detection method based on machine learning

Номер: CN116500456A
Автор: HAN GENWEI, LI LINFENG, YI SHIHUA, XIE YINGHAI, LI XIANHUAI, JING ZEAN, CHEN YONG, ZHAO CHENYANG, LI JUN, ZHOU YU, YU WEIFENG, JIANG XINWEI
Принадлежит:

The invention discloses a ternary lithium battery capacity detection method based on machine learning, and the method comprises the following steps: obtaining battery data, and constructing a data set based on the battery data; performing data dimension reduction and standardization processing on the data set; constructing a battery capacity detection model, training the battery capacity detection model through the data set, and evaluating the trained model; and inputting newly collected battery data into the evaluated battery capacity detection model to realize real-time online detection of the battery capacity. According to the invention, the technical problem of how to accurately and rapidly carry out accurate nondestructive detection on the battery capacity and output the result is solved.

Подробнее
Номер записи: 4
18-07-2023 дата публикации

Fast and accurate rule selection for interpretable decision sets

Номер: US0011704591B2
Автор: Sunny Dhamnani, Dhruv Singal, Ritwik Sinha
Принадлежит: ADOBE INC., Adobe Inc.

An IDS generator determines multiple classes for electronic data items. The IDS generator determines, for each class, a class-specific candidate ruleset. The IDS generator performs a differential analysis of each class-specific candidate ruleset. The differential analysis is based on differences between result values of a scoring objective function. In some cases, the differential analysis determines at least one of the differences based on additional data structures, such as an augmented frequent-pattern tree. A probability function based on the differences is compared to a threshold probability At least one testing ruleset is modified based on the comparison. The IDS generator determines, for each class, a class-specific optimized ruleset based on the differential analysis of each class-specific candidate ruleset. The IDS generator creates an optimized interpretable decision set based on combined class-specific optimized rulesets for the multiple classes.

Подробнее
Номер записи: 5
18-07-2023 дата публикации

Wind power prediction method based on proxy-assisted evolutionary neural network structure search

Номер: CN116451049A
Автор: JIN HUAIPING, ZHANG KEHAO, JIN HUAIKANG, WANG BIN, YANG BIAO, QIAN BIN
Принадлежит:

The invention discloses a wind power prediction method based on proxy-assisted evolutionary neural network structure search. Aiming at the problems that the structure and delay variables of the deep neural network are difficult to select and the existing evolutionary neural network structure search calculation efficiency is low, the fitness is indirectly evaluated through the proxy auxiliary model so as to improve the search efficiency. The method comprises the following specific steps: firstly, capturing high-dimensional nonlinear spatial-temporal characteristics by using a deep neural network, and meanwhile, bringing a delay variable into an evolutionary neural network structure for searching; secondly, performing joint coding on the deep neural network structure and a delay variable; then improving the search efficiency of evolutionary neural network structure search by using an agent model to obtain an optimal neural network structure; and finally, carrying out retraining optimization ...

Подробнее
Номер записи: 6
25-08-2023 дата публикации

Multi-distribution-center electric vehicle path planning method

Номер: CN116644874A
Автор: CHENG MING, WEI XIAOJIE, ZHU RUIFAN, XU WEI, XU HUI, LI JIN, JIANG YONGDONG, XIE GANG, LI YITONG
Принадлежит:

The invention relates to the technical field of logistics distribution, in particular to a multi-distribution-center electric vehicle path planning method. According to the invention, modeling is carried out on the problem of distribution with multiple distribution centers, multiple client points and multiple electric vehicles, and the problem is solved by using a genetic algorithm taking a machine learning model as an evaluation module. Meanwhile, in the genetic algorithm, key genes of chromosomes in the genetic algorithm are judged according to evaluation of a machine learning model, iteration of the genetic algorithm is guided to a certain extent through the key genes, and the solving efficiency of the genetic algorithm on the problem is improved. The method effectively solves the problems of multiple distribution centers of the electric vehicle in a city and the route of considering the endurance, reduces the electric charge cost of the electric vehicle during distribution, and reduces ...

Подробнее
Номер записи: 7
25-08-2023 дата публикации

Vehicle-mounted network intrusion detection method based on stacking integration algorithm

Номер: CN116647844A
Автор: SHANG WENLI, TU JIANGJIAN, JIE HAI, CAO ZHONG, DING LEI, ZHANG MAN, HUAN SHA, SHI HAOTIAN
Принадлежит:

The invention discloses a vehicle-mounted network intrusion detection method based on a stacking integration algorithm. The method comprises the following steps: acquiring a network data set of an external network; carrying out data preprocessing on the acquired network data of the vehicle external network; performing feature engineering on the vehicle external network data after data preprocessing; carrying out model training through three tree-based integration algorithms; and detecting the network data of the extranet of the vehicle to perform normal and abnormal classification, and outputting a processing result. According to the method, a large-scale data set can be processed, meanwhile, the training and prediction speed is high, and the requirement of real-time intrusion detection can be met.

Подробнее
Номер записи: 8
28-07-2023 дата публикации

Urban health risk space identification optimization method, system and device and storage medium

Номер: CN116502910A
Автор: YAN JIAQI, CHEN RAN, WEI ZONGCAI, MA HAO, ZHEN WENZHI, CHEN TONG, LI CHONG
Принадлежит:

The invention discloses an urban health risk space identification optimization method and system, computer equipment and a storage medium, and the method comprises the steps: obtaining basic geographic information data and network open source data of a target city, forming a first sample data set, and carrying out the preprocessing to obtain a second sample data set; performing statistics and processing on the data in the second sample data set to form health risk space related indexes, obtaining a third sample data set, constructing a random forest model, training the random forest model, obtaining a health risk space identification model, performing health risk identification on the grid of the urban to-be-detected region, extracting related data of a high-risk region, and obtaining a health risk space identification result. Forming a fourth sample data set; optimizing and adjusting part of morphological indexes in the fourth sample data set to obtain a fifth sample data set; and according ...

Подробнее
Номер записи: 9
28-07-2023 дата публикации

Viewpoint extraction method, electronic equipment and computer readable storage medium

Номер: CN116501880A
Автор: ZHANG JIADONG, XU XIANCAI, XIAO RONGCHANG, XIONG LEI
Принадлежит:

The invention discloses a viewpoint extraction method, electronic equipment and a computer readable storage medium. The method comprises the following steps: acquiring an extraction result after positive sample training data is processed by an information extraction model in the viewpoint extraction device; according to the extraction result, generating training negative samples of the aspect word classification model, the sentiment classification model and the category and sentiment pairing recognition model in the viewpoint extraction device; according to the positive sample training data and the generated training negative sample, training the aspect word classification model, the sentiment classification model and the category and sentiment pairing recognition model in the viewpoint extraction device; and processing a to-be-extracted text by using the trained viewpoint extraction device to obtain viewpoint information in the to-be-extracted text. According to the scheme provided by ...

Подробнее
Номер записи: 10
18-08-2023 дата публикации

Cable charge test data storage method

Номер: CN116611117A
Автор: HOU JIANJIAN, CHEN CHEN, WANG CHANJUAN, SONG JUNHUA, LIU XUEMIN, HE WENJUN, LI YULIN, LIU JUN, ZHANG TAO, HU WEIYING
Принадлежит:

The invention relates to the field of electric digital data processing, and discloses a cable charge test data storage method, which comprises the following steps of: sequentially acquiring cable charge test data according to a test period; preprocessing the cable charge test data, and associating and adding labels to the preprocessed cable charge test data to obtain cable charge label information; inputting the cable charge label information into a pre-trained data storage model to output a storage position node of the cable charge test data; uploading the cable charge test data to a corresponding distributed data storage end according to the storage position node; encrypting the cable charge test data in the distributed data storage end to obtain encrypted data corresponding to the cable charge test data; according to the technical scheme provided by the invention, data storage of the cable charge test data is realized, and the safety and reliability of long-term storage of the cable ...

Подробнее
Номер записи: 11
29-08-2023 дата публикации

Electric nail gun fault identification method based on historical database model analysis

Номер: CN116662890A
Автор: LU GUOFENG
Принадлежит:

The invention relates to the field of electric tools, and discloses an electric nail gun fault identification method based on historical database model analysis, which comprises the steps of collecting and sorting data of an electric nail gun in a working state, and constructing a historical database model; based on the data in the working state in the historical database model, training a machine learning model, and identifying the working state of the electric nail gun; monitoring real-time working data of the electric nail gun, and feeding the real-time working data of the electric nail gun into the machine learning model; according to the real-time working state, evaluated by the machine learning model, of the electric nail gun, a report system is triggered, and a result is fed back to an operator; and storing the real-time working state of the electric nail gun and data corresponding to the real-time working state into a log file. The method is high in intelligent degree, does not ...

Подробнее
Номер записи: 12
14-07-2023 дата публикации

Code classification method based on Hash algorithm

Номер: CN116432125A
Автор: WU WEI, JIANG MI, LUO CHUAN, ZHONG PING
Принадлежит:

The invention discloses a code classification method based on a Hash algorithm. The code classification method comprises the steps that existing code data and corresponding classification results are obtained, and corresponding abstract syntax trees are extracted; calculating to obtain the expression of each abstract syntax tree; constructing a kernel matrix based on the similarity between the abstract syntax trees; constructing an abstract syntax tree classification preliminary model and training to obtain an abstract syntax tree classification model; and carrying out actual code classification by adopting an abstract syntax tree classification model. Complex mathematical calculation and massive parameter learning in a neural network are avoided, meanwhile, the method does not depend on expensive high-end hardware any more, time expenditure can be obviously reduced on the premise that code classification precision is met, the method is especially suitable for code classification tasks ...

Подробнее
Номер записи: 13
17-08-2023 дата публикации

ADVANCED DATA COLLECTION BLOCK IDENTIFICATION

Номер: US20230259586A1
Автор: Martynas JURAVICIUS, Andrius KUKSTA
Принадлежит: OXYLABS, UAB

Systems and methods that allow examination of response data collected from content providers and provide for classification and routing according to the classification. The process of classification employs an unsupervised, or partially unsupervised, Machine Learning classifier model for identifying data collection responses that contains no data, mangled data, or a block, for assigning a classification correspondingly and for feeding the classification decision back to a data collection platform.

Подробнее
Номер записи: 14
18-04-2023 дата публикации

System and method for monitoring touch-screen gestures of users and for user authentication

Номер: US0011630893B2
Автор: Alesis Novik, Andrius Sutas
Принадлежит: BIOCATCH LTD.

The present invention relates to an improved method of providing identification of a user or authentication of a user's identity. More particularly, the present invention relates to an improved method of providing identification of a user or authentication of a user's identity using conditional behavioural biometrics. The present invention seeks to provide an enhanced method of authenticating and/or identifying a user identity using conditional behavioural biometrics. According to a first aspect of the present invention, there is provided a method of generating a user profile for use in identifying and/or authenticating a user on a device, the device equipped with one or more sensors, the method comprising: generating a set of data points from sensory data collected by the one or more sensors; clustering the set of data points to produce a set of data clusters; developing a first classifier for the data clusters, the first classifier being operable to assign a further data point derived ...

Подробнее
Номер записи: 15
15-08-2023 дата публикации

Machine-learning based personalization

Номер: US0011727082B2
Автор: Aleksandr A. Orlov, Tetiana Kostenko
Принадлежит: SITECORE CORPORATION A/S, Sitecore Corporation A/S

A system, method, and apparatus provide the ability to generate and deliver personalized digital content. Multiple content tests are performed by presenting different variants of content to a set of different consumers of one or more consumers. A machine learning (ML model is generated and trained based on an analysis of results of the multiple content tests. Based on the ML model, personalization rules, that specify a certain variance for a defined set of facts, are output. The personalization rules are exposed to an administrative user who selects one or more of the personalization rules. A request for content is received from a requesting consumer. Based on similarities between the defined set of facts and the requesting consumer, a subset of the selected personalization rules are selected. The content is personalized and delivered to the requesting consumer based on the further selected personalization rules.

Подробнее
Номер записи: 16
24-05-2023 дата публикации

IMAGE SEGMENTATION

Номер: EP4184451A1
Автор: REDFORD, John, SAMANGOOEI, Sina
Принадлежит:

In one aspect, hierarchical image segmentation is applied to an image formed of a plurality of pixels, by classifying the pixels according to a hierarchical classification scheme, in which at least some of those pixels are classified by a parent level classifier in relation to a set of parent classes, each of which is associated with a subset of child classes, and each of those pixels is also classified by at least one child level classifier in relation to one of the subsets of child classes, wherein each of the parent classes corresponds to a category of visible structure, and each of the subset of child classes associated with it corresponds to a different type of visible structure within that category.

Подробнее
Номер записи: 17
30-06-2023 дата публикации

Early warning system for laboratory equipment

Номер: CN116363843A
Автор: ZHANG ZEYU, ZHOU JIAN
Принадлежит:

The invention discloses a laboratory equipment early warning system, comprising a storage module which forms a database based on distributed file storage and stores data in a document mode in an unstructured form; the data processing module is used for carrying out continuous analysis processing and machine learning work on the data in the database, so that the data can be continuously utilized and learned by an early warning algorithm; and the early warning analysis module is used for an early warning algorithm. The reliability and accuracy of equipment early warning can be greatly improved through improvement in three aspects of data storage, data processing and an early warning algorithm, a user can intervene in time, and the service life of the equipment is prolonged.

Подробнее
Номер записи: 18
01-08-2023 дата публикации

Processing method and device for predicting travel destination of user and medium

Номер: CN116522207A
Автор: MA KE
Принадлежит:

The invention provides a user travel destination prediction processing method and device and a medium, and the method comprises the steps: carrying out the prediction processing of an end point of a target user through employing a target prediction model trained based on to-be-trained data according to a starting point and starting time of the target user; wherein the acquisition mode of the to-be-trained data is as follows: according to a frequent visit place FVP set in historical travel trajectory data of the to-be-tested user, solving a total space-time entropy rate corresponding to the to-be-tested user; and judging whether the total space-time entropy rate is smaller than a preset entropy rate, and if the total space-time entropy rate is smaller than the preset entropy rate, taking the historical travel trajectory data of the to-be-tested user as to-be-trained data. Compared with the prior art, the method provided by the invention has the advantages that personalized destination prediction ...

Подробнее
Номер записи: 19
04-08-2023 дата публикации

Botnet traffic detection method based on deep reinforcement learning

Номер: CN116541762A
Автор: ZHAO YUNTAO, FENG YONGXIN, HU YUTAO
Принадлежит:

The invention designs a Botnet traffic detection method based on deep reinforcement learning, and the method comprises the steps: constructing an intelligent OneR-DQN detection model based on an OneR classifier in machine learning and a deep Q network of deep reinforcement learning, and facing an existing Botnet traffic data set; firstly, data sets are combined, data preprocessing is performed, and features capable of being used for classification and training are reserved; secondly, testing and judging various features in the data set one by one by using an OneR classifier, and collecting proper features to be given to a DQN model for training; and finally, continuously extracting independent experiences and training samples by using a special experience pool mechanism of the DQN to carry out cross training, thereby improving the detection accuracy. According to the method, a mixed data set of four CIC data sets of CIC-IDS2017, CIC-DoS2017, CIC-IDS2018 and CIC-DDoS2019 is used, a verification ...

Подробнее
Номер записи: 20
18-07-2023 дата публикации

Financial big data management system based on time sequence neural network

Номер: CN116452335A
Автор: WANG CHAO, WANG JUE, FANG QIUYAN
Принадлежит:

The invention discloses a financial big data management system based on a time sequence neural network. The financial big data management system comprises a financial big data storage module, a financial big data monitoring module and a financial big data prediction module, according to the invention, monitoring and early warning thinking is introduced into a deep learning framework, a financial big data management system is established based on a database, transaction tracking and a sequential neural network, the network structure of the sequential neural network is changed, future transaction risks are analyzed based on transaction data anomaly detection and transaction data prediction, and predictive management is carried out.

Подробнее
Номер записи: 21
01-08-2023 дата публикации

Cardiac shock prognosis prediction and early warning system, equipment and storable medium

Номер: CN116525105A
Автор: SUN YUHUI, HE KUNLUN
Принадлежит:

The invention provides a cardiac shock prognosis prediction and early warning system and device and a storable medium, and relates to the field of intelligent medical treatment. The system comprises an acquisition unit, a feature extraction unit, a diagnosis unit, a second acquisition unit and a decision-making unit, wherein the acquisition unit is used for acquiring genetic information of a peripheral blood mononuclear cell sample of a patient and a classification label about whether ECMO treatment is given or not; a feature extraction unit performs feature extraction on the genetic information to obtain genetic features; the diagnosis unit judges whether the patient is subjected to ECMO treatment or not on the basis of the genetic characteristics; the second acquisition unit is used for acquiring a cardiac shock staging diagnosis result of the patient; a decision unit selects a therapeutic schedule based on the staging diagnosis results. The system developed by the invention predicts ...

Подробнее
Номер записи: 22
11-08-2023 дата публикации

Method for predicting urban inland inundation and ponding based on Bagging-XGB model

Номер: CN116578931A
Автор: SHANG XIAOWEI, CHEN JINGXIN, XU YUFENG, CHEN TONGHE, YUAN CHUANGCHUANG
Принадлежит:

The invention provides an urban inland inundation and ponding prediction method based on a Bagging-XGB model, and belongs to the technical field of prediction methods. The method comprises the following steps: a) establishing a topological relation between a ponding area and a rainfall monitoring station as well as a pump station; b) acquiring ponding and rainfall monitoring data, pump station operation and rainwater drainage pipeline data in the whole city, obtaining ponding area data according to the topological relation in the step a), judging whether rainfall causes ponding, and establishing model training data; c) balancing data distribution, and dividing the data into training set data and verification set data; d) training a Bagging-XGB model by using the data of the training set; and e) verifying the accuracy and recall rate of the model by using verification set data, and evaluating the generalization ability of the model. According to the method, the waterlogging and ponding conditions ...

Подробнее
Номер записи: 23
11-08-2023 дата публикации

Intersection classification method and apparatus, and electronic device

Номер: CN116578930A
Автор: XIE YONGBING
Принадлежит:

The embodiment of the invention discloses an intersection classification method and device and electronic equipment. The intersection classification method comprises the steps that road sections entering and exiting a target intersection are acquired; determining an azimuth area of the road section at the target intersection; determining the azimuth order of the target intersection according to the azimuth area of the road section at the target intersection; and determining the category of the target intersection according to the azimuth area of the road section at the target intersection and the azimuth order of the target intersection. According to the technical scheme, the implementation is simple, and the category to which a certain intersection belongs can be quickly and accurately determined, so that the quality of high-precision map data is improved, the processing efficiency and accuracy of the high-precision map data are improved, and the use experience of a user is improved.

Подробнее
Номер записи: 24
24-08-2023 дата публикации

NEURAL TREES

Номер: US20230267381A1
Автор: Aditya Vithal NORI, Antonio CRIMINISI, Ryutaro TANNO
Принадлежит:

A predictor has a memory which stores at least one example for which an associated outcome is not known. The memory stores at least one decision tree comprising a plurality of nodes connected by edges, the nodes comprising a root node, internal nodes and leaf nodes. Individual ones of the nodes and individual ones of the edges each have an assigned module, comprising parameterized, differentiable operations, such that for each of the internal nodes the module computes a binary outcome for selecting a child node of the internal node. The predictor has a processor configured to compute the prediction by processing the example using a plurality of the differentiable operations selected according to a path through the tree from the root node to a leaf node.

Подробнее
Номер записи: 25
30-06-2023 дата публикации

Detection tool generation method and device, target detection method and device, electronic equipment and medium

Номер: CN116361653A
Автор: GU CHAO, XU XIAOYONG, QIU SHIHAO, ZHANG HUI, WANG CHANGDONG
Принадлежит:

The invention discloses a detection tool generation method and device, a target detection method and device, electronic equipment and a medium, the method and device are applied to the electronic equipment, and specifically, a radar sensor arranged on a test carrier is used for carrying out target recognition on a target area to obtain a moving target; calculating the sensor speed of the radar sensor according to the carrier speed of the test carrier; performing feature extraction on the parameters of the moving target and the speed of the sensor to obtain a feature set; and performing model training based on the feature set to obtain a target detection tool. When the target detection tool is applied to the radar on the vehicle, due to the fact that the tool can distinguish the ghosting target, the ghosting target can be discarded, target false alarm is avoided, and the reliability of target detection by the radar is improved.

Подробнее
Номер записи: 26
18-07-2023 дата публикации

Citrus fruit quality digital model construction method, system, equipment and terminal

Номер: CN116451473A
Автор: HUANG BEI, WANG PENG, WU BIZHI, WEN MINGXIA, WU SHAOHUI, JIN LONGFEI, LIU FENG, YU XIAOFENG
Принадлежит:

The invention belongs to the technical field of digital model construction, and discloses a citrus fruit quality digital model construction method, system, equipment and terminal, and the method comprises the steps: collecting soil, leaf and fruit samples, and determining detection indexes; preprocessing the data by using maximum value and minimum value normalization, and establishing a relation model between the TSS and soil and leaf nutrition; and constructing a quantitative model of TSS stepwise regression and verifying the credibility of the model. According to the method, the multiple models are fitted, influence factors of the multiple models are sorted, the problem of overfitting of a traditional regression method is effectively solved, and the generalization ability of results is improved on the model level. According to the method, contribution sequencing is carried out on various factors influencing the fruit quality by innovatively utilizing an SHAP analysis method, the contribution ...

Подробнее
Номер записи: 27
04-07-2023 дата публикации

Power terminal equipment fingerprint generation method and device, electronic equipment and storage medium

Номер: CN116389420A
Автор: LI NIGE, SHAO ZHIPENG, LU ZI'ANG, ZHANG TAO, MA YUANYUAN, CHEN LU, WANG TENGYAN, LYU ZHUO, LIU XIN, DOU YUFEI
Принадлежит:

The invention discloses a power terminal equipment fingerprint generation method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring network data traffic of communication with a server side when power terminal equipment is accessed; performing data preprocessing on the network data traffic based on an IP address of a data packet in the network data traffic to obtain a preprocessing result; constructing a Huffman tree based on a preprocessing result; and generating a fingerprint of the power terminal equipment based on a pruning result of the Huffman tree. According to the method and the device, the fingerprint features of the power terminal equipment are directly extracted from the network data traffic, so that the method and the device are more suitable for a real network environment with huge network data traffic; the original network traffic is divided into the network traffic segments based on the IP address, so that fine-grained ...

Подробнее
Номер записи: 28
25-07-2023 дата публикации

Wind turbine generator health degree evaluation and maintenance method, system, equipment and medium

Номер: CN116480534A
Автор: SHA DESHENG, LIU XIAOBO, LI QIAN, ZHANG QING, ZOU XIN
Принадлежит:

The invention discloses a wind turbine generator health degree assessment and maintenance method, system, equipment and medium, which fully integrates the existing information resources by intensively acquiring and effectively utilizing the operation data and multi-dimensional monitoring data of a data medium, breaks the information isolated island and improves the working efficiency. After abnormal value cleaning and missing value reconstruction processing, SCADA and CMS are combined, so that the accuracy of fault diagnosis is improved, the accuracy of health degree evaluation of the wind turbine generator is improved, the monitoring difficulty is reduced, a fault case library can be established according to fault diagnosis data, a health degree evaluation result can be generated, and fault positioning is completed; the functions of the data center are optimized, and the health degree state monitoring and fault diagnosis functions of the wind turbine generator are perfected; through combination ...

Подробнее
Номер записи: 29
01-08-2023 дата публикации

Risk customer detection method and device, storage medium and equipment

Номер: CN116523303A
Автор: WEN GUANGMING, TIAN OU, ZHU YUNMING, CHEN MINGZHONG, LIU WENTAO, CHEN XIN, YI GUO
Принадлежит:

The embodiment of the invention provides a risk customer detection method and device, a storage medium and equipment, and the method comprises the steps: introducing an external data source to detect customer transaction behavior data, customer identity data and equipment use data of a to-be-detected customer on the basis of a blacklist, achieving the in-event interception of a part of transactions, and then carrying out the detection of the customer transaction behavior data, the customer identity data and the equipment use data of the to-be-detected customer. According to the method, the trained risk transaction identification model is used for further detection for the clients which are not intercepted, and the risk transaction identification model is obtained by taking the plurality of transaction data with the labels as training samples, so that whether the clients to be detected are risk clients or not can be determined based on the output of the risk transaction identification model ...

Подробнее
Номер записи: 30
22-08-2023 дата публикации

Method for repairing missing underground water level monitoring data

Номер: CN116627953A
Автор: SUN YONGHUA, ZHANG WANGKUAN, CHENG XINGLU, CAO XUYUE, WANG YANZHAO
Принадлежит:

The invention discloses a method for repairing missing underground water level monitoring data. The method comprises the following steps: step 1, acquiring and arranging the underground water level monitoring data to form an underground water spatio-temporal data set; step 2, performing interpolation on the time sequence data containing missing values by using a BTF model; 3, performing abnormal value detection on the interpolated time sequence data by using an isolated forest model; and step 4, carrying out interpolation on the time sequence data again until the data is completely restored. According to the method, the defect that part of abnormal values are generated when only a BTF interpolation method is adopted is effectively overcome, and the reliability and the accuracy of an interpolation result are improved; the method is suitable for the condition of long-time sequence missing of underground water monitoring data, solves the problems that an existing method is poor in interpolation ...

Подробнее
Номер записи: 31
08-08-2023 дата публикации

Short-term load prediction method and system based on combined model

Номер: CN116562559A
Автор: PENG JUAN, LIN JINGYAN, YU ZHIHAO
Принадлежит:

The invention relates to an electrical load prediction method provided by the embodiment of the invention, and the method comprises the steps: obtaining historical power consumption data, and dividing the historical power consumption data into a plurality of groups of classification data through descriptive statistical analysis and density clustering; multiple groups of model features are constructed based on the classification data, and the multiple groups of model features comprise basic features and additional features obtained through time sequence model prediction; and based on the multiple groups of model features, training the ensemble learning tree model to obtain a power consumption prediction model, and performing power consumption load prediction through the power consumption prediction model. In the scheme of the invention, after users are accurately classified according to data expressions, additional features of a plurality of models are combined, the ensemble learning tree ...

Подробнее
Номер записи: 32
19-10-2023 дата публикации

DISTRIBUTED AND REDUNDANT MACHINE LEARNING QUALITY MANAGEMENT

Номер: US20230334580A1
Автор: Alain Charles Briancon, Jean Joseph Belanger, Chris Michael Coovrey, Travis Stanton Penn, Divya Karumuri, Valisis Sotiris
Принадлежит:

Provided is a process including: writing modelling-object classes using object-oriented modelling of the modelling methods, the modelling-object classes being members of a set of class libraries; writing quality-management classes using object-oriented modelling of quality management, the quality-management classes being members of the set of class libraries; scanning modelling-object classes in the set of class libraries to determine modelling-object class definition information; scanning quality-management classes in the set of class libraries to determine quality-management class definition information; using the modelling-object class definition information and the quality-management class definition information to produce object manipulation functions that allow a quality management system to access methods and attributes of modelling-object classes to manipulate objects of the modelling-object classes; and using the modelling-object class definition information and the quality-management ...

Подробнее
Номер записи: 33
23-06-2023 дата публикации

Prediction model training method and device, computer equipment and storage medium

Номер: CN116306985A
Автор: ZHOU KUI, HUANGFU XIAOJIE, ZHANG QIANNI, WANG HANG
Принадлежит:

The invention relates to a prediction model training method and device, computer equipment and a storage medium. The method comprises the following steps: acquiring historical feature sample data; determining a change information classification type corresponding to each information type classification model in the to-be-trained resource object change information prediction model according to the historical feature sample data; inputting the historical feature sample data into each information type classification model to obtain dichotomy confidence information corresponding to each information type classification model; and according to the dichotomy confidence degree information corresponding to each information type classification model, training the corresponding information type classification model until each dichotomy confidence degree information satisfies a preset condition, and obtaining a trained resource object change information prediction model. By adopting the method, the ...

Подробнее
Номер записи: 34
30-06-2023 дата публикации

Financial product recommendation method and device and storage medium

Номер: CN116362895A
Автор: HUANG JINGCHAO, ZENG WEI, CHEN LINGXIAO, RONG DAFENG
Принадлежит:

The invention discloses a financial product recommendation method and device and a storage medium. The method comprises the steps of obtaining first account feature data of a target account; based on the first account feature data, a random forest financial management prediction model is adopted to obtain financial management tags output by N random trees respectively and tag scores corresponding to the financial management tags output by the N random trees respectively; obtaining financial management label scores corresponding to K financial management labels corresponding to the target account according to the financial management labels output by the N random trees and the label scores corresponding to the financial management labels output by the N random trees; and based on the financial management tag scores corresponding to the K financial management tags, determining a financial management product recommendation result of the target account. According to the method and the device ...

Подробнее
Номер записи: 35
08-08-2023 дата публикации

Model training method, purchase insurance rate prediction method, device, equipment and medium

Номер: CN116561678A
Автор: SONG YIMIN, CAI CHAOCHAO, ZHANG PENG, LI DATING, NIU ZIRU, JIANG HAITAO
Принадлежит:

The embodiment of the invention discloses a model training method, a purchase insurance rate prediction method and device, equipment and a medium, and belongs to the technical field of federal learning. According to the model training method, under the condition that the first electronic equipment and the second electronic equipment are provided with the tags for the first sample, the target features can be screened based on the tags of the two parties, and the tag data of the two parties are fully utilized, so that the training effect of the model can be improved; and the accuracy of the prediction result can be improved when the purchase rate of the user is predicted by using the trained longitudinal federal model subsequently.

Подробнее
Номер записи: 36
01-08-2023 дата публикации

Remote power failure and recovery fault automatic diagnosis method

Номер: CN116522111A
Автор: CUI CHAOYI, ZHANG GANG, REN WEI, ZHANG KAOSHE, FENG SHUANG, XIE TUO, SHI NING, JIA TIANSHENG, CAO JINGHANG, QU LINJING, WANG HAIPENG, ZHANG LAN
Принадлежит:

The invention discloses a remote power failure and recovery fault automatic diagnosis method. The method is specifically implemented according to the following steps: step 1, sampling and analyzing different power failure and recovery fault characteristics; 2, a classification decision tree power failure and recovery fault traceability model; and step 3, decision optimization and power failure and recovery fault reason classification. According to the invention, the blank in the aspect of automatic diagnosis of power failure and recovery failure reasons is filled, and the intelligent level of remote power failure and recovery business operation is improved.

Подробнее
Номер записи: 37
21-07-2023 дата публикации

Depression recognition system based on electroencephalogram-voice bimodal decision fusion

Номер: CN116467672A
Автор: CHEN WANZHONG, GONG LINLIN
Принадлежит:

The invention discloses a depression identification method based on electroencephalogram-voice bimodal decision fusion, and belongs to the technical field of digital signal processing. The invention aims to provide a depression recognition system based on electroencephalogram-voice bimodal decision fusion, which adopts a D-S evidence theory decision fusion strategy combining a capsule graph neural network and a Transform network, and aims to improve the recognition and classification objectivity and accuracy of a detection device on depression crowds. The method comprises the steps of collecting electroencephalogram and voice signals, preprocessing data, extracting features of electroencephalogram and voice, constructing and training a capsule graph neural network and a Transform network, and classifying and recognizing depression of a test set. According to the method, the problems of large individual difference influence and low depression classification identification degree caused by ...

Подробнее
Номер записи: 38
04-07-2023 дата публикации

Method and system for deploying power supply units in area, electronic equipment and storage medium

Номер: CN116388167A
Автор: MA PEIRUI
Принадлежит:

The invention discloses a method for deploying power supply units in a region, belongs to the technical field of power supply distribution, and solves the problem that the distribution of the power supply units in the region is unbalanced due to the fact that the saturation load density and the saturation net rack load of a land parcel cannot be predicted and calculated according to the space load of the land parcel in an existing method. Performing cleaning and noise reduction processing on the plot planning data set to obtain a noise reduction data set; constructing a power supply unit deployment model through a random forest algorithm, and training, testing and verifying the power supply unit deployment model; predicting and calculating land saturation load density and saturation grid load based on the power supply unit deployment model; according to the method and the device, by acquiring the plot planning information in the region, the later power supply unit deployment and calculation ...

Подробнее
Номер записи: 39
23-11-2023 дата публикации

SYSTEMS AND METHODS FOR GENERATING GRADIENT-BOOSTED MODELS WITH IMPROVED FAIRNESS

Номер: US20230377037A1
Автор: Sean Javad Kamkar, Geoffrey Michael Ward, Jerome Louis Budzik
Принадлежит:

Systems and methods for generating tree-based models with improved fairness are disclosed. The disclosed process generates a first tree-based machine learning model, which is preferably trained to predict if a financial loan will be repaid. The process also determines an accuracy of the first tree-based machine learning mode. In addition, the process determines a fairness of the first tree-based machine learning model. The fairness is preferably associated with at least one of gender, race, ethnicity, age, marital status, military status, sexual orientation, and disability status. The process then generates a second different tree-based machine learning model, which is preferably trained based on the accuracy of the first tree-based machine learning model and the fairness of the first tree-based machine learning model. The process then combines the first tree-based machine learning model and the second tree-based machine learning model to produce a gradient-boosted machine learning model ...

Подробнее
Номер записи: 40
18-07-2023 дата публикации

Automatic identification technology for fake-licensed ship

Номер: CN116451154A
Автор: ZHENG HAILIN, HU QINYOU, YANG CHUN, CHEN PU
Принадлежит:

The invention discloses a fake-licensed ship automatic identification technology. In order to overcome the problem that point adjustment and other track abnormal conditions existing in ship moving track identification cannot be solved in the prior art, a local threshold value window and a mean value filtering window are designed, a local speed threshold value and a global speed threshold value are calculated, three types of abnormal track points including isolated abnormal points, continuous abnormal points and obvious identification abnormal points are found, and the abnormal track points are identified. Respectively corresponding to track points, noise points and confusion points of the fake-licensed ship; the abnormal track points are screened out by counting the speed distribution of the track points of the ship, calculating the bow deflection rate between the track points, setting the speed threshold of the track points according to the speed distribution probability of the track points ...

Подробнее
Номер записи: 41
02-05-2023 дата публикации

POI classification and model training method and device, equipment and storage medium

Номер: CN116049335A
Автор: ZHANG XIN
Принадлежит:

The invention provides a POI classification and model training method and device, equipment and a storage medium, and relates to the technical field of data processing, in particular to the technical field of electronic maps, big data and the like. The POI classification method comprises the following steps: if a first POI does not meet a preset rule and a second POI associated with the first POI does not exist in target data, obtaining attribute data of the first POI; wherein the first POI is a POI to be classified in map data, and the target data is different from the map data; obtaining a POI classification model, wherein the POI classification model is trained based on sample POIs in the target data; and processing the attribute data of the first POI by adopting the POI classification model so as to determine a target category of the first POI. According to the method and the device, the POI classification accuracy can be improved.

Подробнее
Номер записи: 42
18-07-2023 дата публикации

Cement raw material mill raw material fineness index prediction method, model training method and system

Номер: CN116451856A
Автор: LI WANYING, HU YAOLIN, JING SHIQING, OUYANG BAOQING, JIANG KAIFANG, WANG GUOXUN, LIU YUTONG
Принадлежит:

The invention provides a cement raw material mill raw material fineness index prediction method and a model training method and system, and belongs to the technical field of cement production. The method comprises the following steps: obtaining target process parameters in a cement raw material grinding raw material production process; calculating statistical characteristics of the target process parameters to obtain statistical characteristic values used for describing data characteristics of the target process parameters; inputting the target process parameters into a pre-trained auto-encoder, and extracting an output result of an encoder in the auto-encoder to obtain hidden layer information; and obtaining process parameter index information according to the hidden layer information and the statistical characteristic value, and inputting the process parameter index information into a pre-trained fineness index prediction model to obtain a fineness index prediction result. According to ...

Подробнее
Номер записи: 43
29-08-2023 дата публикации

Target grouping method and device based on confidence decision tree clustering

Номер: CN116662830A
Автор: JIAO LIANMENG, YANG HAOYU, WANG JIANING, GOU SHENGJIE, LIU ZHUN, LIANG YAN, PAN QUAN
Принадлежит:

The invention provides a target grouping method and device based on a confidence decision tree. The method comprises the following steps: obtaining individual attributes, capability attributes and communication relationships of targets through detection; taking all targets as root nodes; performing node division from a root node; traversing all the nodes to be divided until all the nodes do not need to be divided, and finally forming a confidence decision tree; and determining the group of the target according to the confidence of the target belonging to a certain sub-node in the confidence decision tree. According to the method, the cutting points during node division of the confidence decision tree are determined in the division process, and the confidence division method based on the cutting points is provided, so that child nodes are generated, and finally the confidence decision tree is formed. And finally, an adversarial group obtained by the confidence decision tree can be adjusted ...

Подробнее
Номер записи: 44
04-07-2023 дата публикации

Ore-bearing pegmatite ground spectrum identification method based on full spectrum section

Номер: CN116380809A
Автор: ZHANG NANNAN, ZHAO TONGYANG, CHEN LI, ZHANG HAO, TAO JINTAO, LIAO SHIBIN, CUI SHICHAO, CHANG JINYU, CHENG WANWEN
Принадлежит:

The invention discloses an ore-bearing pegmatite ground spectrum identification method based on a full spectrum band, and provides a method for identifying ore-bearing pegmatite in ground spectrum data based on a mode of combining random forest and principal component transformation data enhancement in order to solve the problem of accurate identification of ore-bearing pegmatite. The method comprises the following steps: 1, rock sample collection, ground spectrum measurement and ground spectrum pretreatment; step 2, resampling the full-spectrum spectral data to determine the optimal spectral resolution; step 3, using two feature selection methods of minimum angle fitting and genetic algorithm to carry out selection of intelligent feature subsets; and 4, inputting a random forest recognition model through data enhancement processing, and finally judging an optimal recognition model based on an AUC value. The method disclosed by the invention can reveal the implicit characteristics in the ...

Подробнее
Номер записи: 45
15-08-2023 дата публикации

Multi-classification prediction model training method, performance assessment score prediction method and device

Номер: CN116595362A
Автор: WU YANGE, WEN JIANBO, LEI ZHUOMIN, CHEN JING
Принадлежит:

The invention provides a multi-classification prediction model training method, relates to the technical field of artificial intelligence, and can be applied to the technical field of finance. The method comprises the following steps: inputting a personnel characteristic data sample into a rough classification initial prediction model to obtain an initial classification result of the personnel characteristic data sample; determining a fusion sample according to the initial classification result and the personnel feature data sample; inputting the fusion sample into a fine classification initial prediction model to obtain a target classification result of the fusion sample; adjusting parameters of the initial prediction model by adopting a matrix partition dimensionality reduction Bayesian tuning method so as to determine an optimal hyper-parameter combination; and keeping the optimal hyper-parameter combination to determine a target multi-classification prediction model. The invention further ...

Подробнее
Номер записи: 46
28-07-2023 дата публикации

Key signal detection system and method

Номер: CN116502071A
Автор: CHANG XING
Принадлежит:

The invention relates to the technical field of satellite communication systems, and discloses a key signal detection system and method, and the system comprises a satellite signal collection module; a satellite signal feature analysis module; a key signal detection module; wherein the satellite signal acquisition module acquires historical satellite signals and satellite signals to be detected; a satellite signal feature analysis module extracts a satellite signal feature matrix; and the key signal detection module performs deep reinforcement learning model training by using the satellite signal feature matrix of the satellite signal to be detected, and performs decision classification tree training by using the satellite signal feature matrix of the historical satellite signal so as to detect whether the satellite signal has the appearance or disappearance of the key signal. According to the method, deep reinforcement learning model training and decision classification tree training are ...

Подробнее
Номер записи: 47
29-08-2023 дата публикации

Spinal fracture detection in x-ray images

Номер: US0011741694B2
Автор: Larissa Christina Schudlo
Принадлежит: MERATIVE US L.P.

Methods and systems for detecting a vertebral fracture within an x-ray. One method includes receiving a chest x-ray image and identifying a plurality of vertebrae represented in the chest x-ray image. The method further includes extracting a plurality of image patches from the chest x-ray image, each image patch of the plurality of image patches including a portion of the chest x-ray image representing one of the plurality of vertebrae identified in the chest x-ray image. The method further includes sequencing the plurality of image patches into an ordered sequence of image patches, and assigning, with a deep learning model applied to the ordered sequence of image patches, a classification to each of the plurality of image patches indicating whether the image patch represents a fractured vertebra or an unfractured vertebra.

Подробнее
Номер записи: 48
01-08-2023 дата публикации

Machine learning-based root cause analysis of process cycle images

Номер: US0011715200B2
Автор: Naghmeh Rezaei, Pedro Miguel Filipe Cruz
Принадлежит: Illumina, Inc.

The technology disclosed relates to classification of process cycle images to predict success or failure of process cycles. The technology disclosed includes capturing and processing images of sections arranged on an image generating chip in genotyping process. Image description features of production cycle images are created and given as input to classifiers. A trained classifier separates successful production images from unsuccessful or failed production images. The failed production images are further classified by a trained root cause classifier into various categories of failure.

Подробнее
Номер записи: 49
19-10-2023 дата публикации

CROSS IN-DATABASE MACHINE LEARNING

Номер: US20230334303A1
Автор: Marco Antonio Carniel Furlanetto, Alessandro Parolin, Cristiano Ruschel Marques Dias, Alejandro Salinger
Принадлежит:

Techniques for implementing cross in-database machine learning are disclosed. In some example embodiments, a computer-implemented method comprises training a machine learning model in a first database instance using a machine learning algorithm and a training dataset in response to receiving a request to train, serializing the trained machine learning model into a binary file in response to the training of the machine learning model, recreating the trained machine learning model in a second database instance using the binary file in response to receiving a request to apply the machine learning model, and generating an inference result by applying the recreated trained machine learning model on the inference dataset in the second database instance.

Подробнее
Номер записи: 50
01-08-2023 дата публикации

Soil moisture downscaling method considering vegetation memory and fine-scale vegetation index

Номер: CN116522090A
Автор: CUI CHANGLU, XIANG DAXIANG, LI ZHE, CHEN ZHE, JIANG YING, CHEN XICHI, ZHAO JING, WU YIBANG, WEN XIONGFEI, LI JINGWEI
Принадлежит:

The invention discloses a soil moisture downscaling method considering vegetation memory and a fine-scale vegetation index. The method comprises the following steps: 1) generating the fine-scale vegetation index with a 30m spatial resolution; 2) acquiring NDVI data having hysteresis with the soil moisture data, and designing vegetation memory data; 3) obtaining MODIS data, SRTM data, SoilGrids data, ERA5-Land data and Landsat data of the target area, and performing space-time matching with the SMAP data; and 4) combining the vegetation memory data, the fine-scale vegetation index and the data obtained in the step 3) to construct a soil moisture downscaling model so as to obtain a high-spatial-resolution soil moisture product. According to the method, high-spatial-resolution soil moisture data can be generated, the result has higher spatial expression ability, detailed information of soil moisture can be reflected, and the application ability of the soil moisture data in a small-scale range ...

Подробнее
Номер записи: 51
29-08-2023 дата публикации

Game content dynamic adjustment method, electronic equipment and storage medium

Номер: CN116650973A
Автор: ZHUANG YUAN, WANG PING, ZHAO MINGLEI
Принадлежит:

The invention provides a game content dynamic adjustment method, electronic equipment and a storage medium, and the method comprises the steps: collecting feature parameters of a user under different game playing methods, the feature parameters comprising a time parameter and a content parameter; calculating tendency weights of the user to different game playing methods according to the time parameters; preprocessing the content parameters, and performing label classification on users to form feature parameters; training different prediction models corresponding to different game playing methods by using the tendency weight and the feature data; and when the player plays a game, predicting the label by using the prediction model, and dynamically adjusting the game content of the player according to a prediction result. According to the scheme, the game time, the game feature parameters and the like of the user are collected and processed to be used for training the prediction model for ...

Подробнее
Номер записи: 52
14-04-2023 дата публикации

Anti-tumor molecule strengthening method based on graph neural network

Номер: CN115966266A
Автор: SHI YUTIAN, WANG BEILUN, JIN QIAO, WANG MINWEN
Принадлежит:

The invention designs an anti-tumor molecule reinforcement learning method based on a graph neural network, and the method comprises the following steps: 1, dividing molecules into anti-tumor (positive) and non-anti-tumor (negative) types according to molecular tags in a database, and 2, inputting an obtained graph into a proposed anti-tumor molecule reinforcement model, 3, applying constraints, carrying out target optimization, ensuring drug-likeness of the molecules in a molecule strengthening process, and obtaining the local structural features of the anti-tumor molecules according to implicit expressions of different property learning graphs of the positive molecules and the negative molecules so as to be used for one-step generation and optimization of the molecules; the method comprises the following steps of 1, obtaining a local molecular structure, 2, substituting the obtained local molecular structure to carry out structure modification and optimization of anti-tumor molecules, ...

Подробнее
Номер записи: 53
25-05-2023 дата публикации

PREDICTIVE USE OF QUANTITATIVE IMAGING

Номер: US20230157644A1
Автор: Glen W. McLaughlin
Принадлежит:

The present disclosure provides systems and methods for predicting a disease state of a subject using ultrasound imaging. The method includes identifying at least one quantitative measurement of a subject using ultrasound imaging, the at least one quantitative measurement included as part of quantitative information of the subject gathered based on the ultrasound imaging, comparing the at least one quantitative measurement to a first predetermined standard to determine a first initial value, the first predetermined standard falling within a first range of quantities, identifying at least one qualitative measurement of the subject using the ultrasound imaging, the at least one qualitative measurement included as part of qualitative information of the subject gathered based on the ultrasound imaging, comparing the at least one qualitative measurement to a second predetermined standard to determine a second initial value, the second predetermined standard falling within a second range of quantities ...

Подробнее
Номер записи: 54
27-06-2023 дата публикации

Encrypted traffic classification method and device, electronic equipment and computer readable medium

Номер: CN116340814A
Автор: WANG XINGWEI, LU XIJIA, WANG XUE, WU DONGKUO, JIA JIE
Принадлежит:

The embodiment of the invention discloses an encrypted traffic classification method and device, electronic equipment and a computer readable medium. A specific embodiment of the method comprises the following steps: acquiring network flow data by using a network packet capturing tool, and storing unprocessed network flow data in a packet format; two types of traffic characteristics are extracted from the unprocessed network traffic data, and the two types of traffic characteristics comprise a statistical characteristic and a time sequence characteristic; processing the statistical features and the time sequence features by using a classifier based on machine learning; and analyzing a processing result by using a two-layer classifier to obtain a classification result of the network flow data. According to the embodiment of the invention, the processing result is analyzed by using two layers of classifiers, and the accuracy of the classification effect of the encrypted traffic is improved ...

Подробнее
Номер записи: 55
14-07-2023 дата публикации

Transformer bushing partial discharge mode identification method based on DOA-RF

Номер: CN116432022A
Автор: HUANG XINBIN, LUO CHUANSHENG, HUANG WEIBIN, ZHOU ZHOU, ZHANG YUNLIANG, HUANG YUANFEI, HUANG XIAOMING, XIAO XINGYUN, YANG PENGCHENG, QIN JUNJIE, LI YAN, LIANG WEIYONG
Принадлежит:

The invention discloses a transformer bushing partial discharge mode identification method based on DOA-RF, and the method comprises the steps: S1, obtaining transformer bushing partial discharge UHF waveform data, extracting UHF feature parameters, and carrying out the normalization processing; s2, carrying out dimensionality reduction on the feature parameters, and constructing a feature parameter vector group; s3, initializing the RF model, setting the value range of the number of decision trees (Nestimators) and the value range of the maximum feature number (maxfeatures), and carrying out random assignment on the value range of the number of decision trees (Nestimators) and the value range of the maximum feature number (maxfeatures); s4, initializing a field dog population, and searching optimal solutions of two parameters including the number of decision trees and the maximum feature number in the RF model through a field dog optimization algorithm DOA; and S5, constructing a fault ...

Подробнее
Номер записи: 56
07-07-2023 дата публикации

Text recognition method and device, computer equipment and storage medium

Номер: CN116402058A
Автор: HUANG XIUWEN, ZHANG BIN, LIU YINGKAI, LIANG SEN
Принадлежит:

The invention relates to the technical field of big data, provides a text recognition method and device, computer equipment, a storage medium and a computer program product, and can be specifically applied to the financial field or other related fields. According to the method, the efficiency and accuracy of obtaining the semantic information of the text can be improved. The method comprises the steps of obtaining a to-be-recognized text; the to-be-recognized text is a business text in a financial system; fusing the to-be-recognized text with the semantic guidance text to obtain a fused text; the semantic guidance text comprises shielding words; the shielding word is used for representing the replacement position of the candidate word; semantic recognition is conducted on the fused text, and the probability that the shielding words in the fused text are replaced with the candidate words is obtained; determining a target word of which the corresponding probability meets a first preset probability ...

Подробнее
Номер записи: 57
11-08-2023 дата публикации

Fault diagnosis method and fault diagnosis device for power system and electronic device

Номер: CN116578869A
Автор: GUO HAIPING, GUO QI, LU YUANHONG, GUO TIANYU, ZHANG JIE, HUANG LIBIN, TU LIANG, HU YUN, HONG ZEQI, LIU YUYAN, SU MINGZHANG, WU WENCONG, CHEN ZHIHAO, HU YUFENG, ZHAO YANJUN
Принадлежит:

The invention provides a fault diagnosis method and device of a power system and an electronic device. The method comprises the steps of performing data preprocessing and feature extraction on a received original data set to obtain a target data set; a model structure and hyper-parameters of a preset fault diagnosis model are determined based on a target algorithm, the preset fault diagnosis model is trained based on the target data set, a target fault diagnosis model is obtained, and the target algorithm comprises one of a genetic algorithm, a Bayesian optimization algorithm and a fault diagnosis algorithm. The preset fault diagnosis model is used for performing fault diagnosis on the target power system; and inputting the power system data of the target power system received in real time into the target fault diagnosis model to obtain a fault diagnosis result of the target power system, thereby solving the problem of low accuracy and efficiency caused by performing fault diagnosis on ...

Подробнее
Номер записи: 58
08-08-2023 дата публикации

Coal supply and demand situation early warning method and system

Номер: CN116562450A
Автор: GONG XIANFU, PENG BO, LI YAODONG, LIU XIAOGUANG, XIAO JIANHUA, GU YANXUN, LIN HAISHENG
Принадлежит:

The invention discloses a coal supply and demand situation early warning method and system, and the method comprises the steps: obtaining historical coal supply and demand data, such as the historical total coal production amount, the total coal consumption amount, the total coal supply amount, the coal price, the electricity-coal inventory, and the number of available days of coal; and inputting historical coal supply and demand data into a coal supply and demand situation early warning model obtained by constructing a decision tree classification model based on a gradient lifting algorithm for calculation to obtain a coal supply and demand early warning result, so that the dispatching center dispatches coal resources according to the coal supply and demand early warning result. The coal supply and demand early warning level is autonomously identified through the gradient lifting classification algorithm, and the coal supply and demand prediction accuracy and prediction efficiency are ...

Подробнее
Номер записи: 59
21-07-2023 дата публикации

Method for establishing random forest model for predicting atmospheric pollution concentration

Номер: CN116467953A
Автор: MA YUE, YIN FEI, ZHANG TAO, WANG WEI, ZHANG YI
Принадлежит:

The invention discloses a method for establishing a random forest model for predicting atmospheric pollution concentration, which comprises the following steps of: 1, scanning a spatial region in sequence, and determining spatial aggregation information of all aggregation regions and non-aggregation regions of atmospheric pollution in the spatial region; 2, determining a relationship type between any two space units; 3, assigning values to the spatial association strength among the corresponding spatial units, and constructing a data-driven spatial weight matrix based on the spatial association strength among all the spatial units; and 4, generating a spatial autocorrelation item by using the data-driven spatial weight matrix, and establishing a random forest model for predicting the atmospheric pollution concentration by taking the spatial autocorrelation item and the common covariable as prediction variables and the atmospheric pollution concentration data of the monitoring station as ...

Подробнее
Номер записи: 60
13-06-2023 дата публикации

Liver cancer occurrence risk prediction model and construction method of network calculator thereof

Номер: CN116259410A
Автор: YAN JUN, NIE GUOLE, ZHONG LINMEI, ZHANG HONGLONG, WANG HAIPING, LI XUN
Принадлежит:

The invention discloses a liver cancer occurrence risk prediction model and a construction method of a network calculator thereof. The construction method comprises the following steps: S1, acquiring clinical and inspection data corresponding to a research object; s2, screening independent prediction features; s3, constructing a prediction model; s4, generating a network calculator; and S5, calculating the prediction probability of the occurrence of the liver cancer. The model for predicting the occurrence risk of the liver cancer related to the family aggregated hepatitis B and a network calculator are integrated into an electronic case system, so that an electronic decision can be provided for a clinician, the clinician can be better helped to evaluate the risk of the liver cancer related to the family aggregated hepatitis B, and therefore, regular follow-up detection can be performed on a patient with high risk of the liver cancer; and early recognition and intervention are carried out ...

Подробнее
Номер записи: 61
23-06-2023 дата публикации

Dynamic graph link generation method

Номер: CN116306780A
Автор: BAO BINGKUN, WANG HONGYU, SHAO XI
Принадлежит:

The invention discloses a dynamic graph link generation method. The method comprises the following steps: acquiring snapshots at different moments; inputting the snapshots at different moments into a preset generative adversarial graph capsule network model to obtain a link between the snapshots at the next moment; wherein the generative adversarial graph capsule network model is obtained by performing double training based on a generative adversarial network, the generative adversarial graph capsule network model is trained by adopting a sample set, and the sample set is obtained based on a topological structure correspondingly generated by each dynamic discrete snapshot graph; the generative adversarial network comprises a generator and a discriminator, and the discriminator is used for discriminating and judging real data and generated data; and the generator is used for generating a topological structure of the snapshot graph at the next moment.

Подробнее
Номер записи: 62
28-07-2023 дата публикации

Big data behavior analysis method and system based on zero-trust network user

Номер: CN116506206A
Автор: CHEN LIANG, SONG YUBO
Принадлежит:

The invention discloses a big data behavior analysis method and system based on zero-trust network users, and the method and system are applied to a server, and the method comprises the steps: extracting user information logging in a terminal device from a big data behavior, carrying out the evaluation of the user information, and generating a first evaluation value; evaluating the operation environment of the terminal equipment according to the terminal equipment information, and generating a second evaluation value; adding the first evaluation value and the second evaluation value to obtain a security evaluation value APG, and obtaining a corresponding interaction behavior data permission generated by the user on the corresponding terminal equipment according to the security evaluation value APG; and obtaining permission data corresponding to the user according to the interaction behavior data permission to construct a behavior-intention bipartite graph model, obtaining intention data ...

Подробнее
Номер записи: 63
29-08-2023 дата публикации

Image segmentation

Номер: US0011741368B2
Автор: John Redford, Sina Samangooei
Принадлежит: Five AI Limited

In one aspect, hierarchical image segmentation is applied to an image formed of a plurality of pixels, by classifying the pixels according to a hierarchical classification scheme, in which at least some of those pixels are classified by a parent level classifier in relation to a set of parent classes, each of which is associated with a subset of child classes, and each of those pixels is also classified by at least one child level classifier in relation to one of the subsets of child classes, wherein each of the parent classes corresponds to a category of visible structure, and each of the subset of child classes associated with it corresponds to a different type of visible structure within that category.

Подробнее
Номер записи: 64
23-06-2023 дата публикации

Data testing method and device and nonvolatile storage medium

Номер: CN116303061A
Автор: LI XUESU
Принадлежит:

The invention discloses a data testing method and device and a nonvolatile storage medium. The method comprises the following steps: acquiring a target test scene of a target data source server; based on the target test scene and a preset test standard, determining a target test case in N predetermined test cases included in a preset initial test case set, N being a positive integer greater than or equal to 1; adopting the target test case to test the operation and maintenance data obtained by synchronization in the target data source server to obtain a case test result; and based on the case test result, updating the initial test case set to obtain a target test case set. Through the method and the device, the problem of non-ideal data testing efficiency in related technologies is solved.

Подробнее
Номер записи: 65
04-08-2023 дата публикации

Power transmission line galloping early warning method, device and equipment and storage medium

Номер: CN116541780A
Автор: LIANG YUN, WANG JINYU, WANG CHAO, LI ZHE, LU MING, ZHANG LU, PANG KAI, LIU SHANFENG, CHEN CEN, GENG JUNCHENG, ZHANG XIAOFEI, ZHAO JIAN, LIU ZEHUI, LI YUQUAN, GENG JINFENG, YANG XIAOHUI
Принадлежит:

The invention provides a power transmission line galloping early warning method, device and equipment and a storage medium, and belongs to the technical field of power transmission line galloping early warning. The early warning method comprises the following steps: constructing a power transmission line galloping early warning model, and carrying out early warning on galloping of a power transmission line by utilizing the early warning model; the early warning model is constructed by using an Adaboost algorithm, and the construction method comprises the steps: collecting line galloping historical data, and forming a training sample set; iteratively generating N decision tree stump weak classifiers; iteratively generating N horse field system weak classifiers; and merging classification results to obtain a strong classifier. Power transmission line galloping early warning can be carried out according to meteorological conditions and line parameters, the algorithm structure is simple, and ...

Подробнее
Номер записи: 66
21-07-2023 дата публикации

Data pre-training and reasoning method and device based on robust optimization Bert model

Номер: CN116468093A
Автор: LIU DONGBAO
Принадлежит:

The invention discloses a data pre-training and reasoning method and device based on a robust optimization Bert model, and belongs to the field of data pre-training and reasoning, and the method comprises the following steps: 1, predicting and filtering data through a self-training technology, and amplifying a generated data sample; 2, annotating and coding the original data by adopting a RoBERT model; 3, performing sample selection by adopting data cleaning and confidence evaluation; and 4, recalling the data samples subjected to error filtering. The device comprises a self-training module, an annotation coding module, a data sample selection module and a data sample recall module. According to the invention, a new self-training selection framework is provided, and two selectors are used for selecting high-quality samples from original data for amplification.

Подробнее
Номер записи: 67
22-08-2023 дата публикации

Decision tree processing for chatbot dialog flow

Номер: US0011736418B2
Автор: Lyudmil Valentinov Pelov
Принадлежит: Oracle International Corporation, ORACLE INTERNATIONAL CORPORATION

Implementations described here provide a chatbot system that is configurable to meet the needs of the user. In one implementation, a chatbot system is configured to utilize and vary decision trees, decision tree complexity, and decision tree selection adaptive to the needs of the end user. Further, implements may utilize decision trees that are configured as static or dynamic using independent leaves and skillsets to allow for variations in the level of sophistication needed for a chatbot conversation. In other implementations, a chatbot system may be configured to assess the chatbot scenario and requirements in order to adapt processing requirements in order to increase or decrease processing threads to vary processing efficiency relative to the needs of the user. In some scenarios, the chatbot system described herein is introspective thereby using feedback and data to be adaptive to chatbot session errors and self-healing.

Подробнее
Номер записи: 68
08-08-2023 дата публикации

Prognostic prediction device for regularization of evolution trend of pseudo tag

Номер: CN116564524A
Автор: HU DANQING, ZHU XIAOFENG, SU HUI
Принадлежит:

The invention discloses a prognosis prediction device for pseudo-tag evolution trend regularization, which comprises the following steps of: acquiring a prognosis data set containing a plurality of patients, selecting a maximum time value of an end point event or visiting loss of a patient, dividing the maximum time value into a plurality of time periods at equal intervals, calculating prognosis tag results of all patients in each time period, and calculating the prognosis tag results of all patients in each time period; wherein a KM curve is used for calculating a time period when a patient is out of visit and a pseudo tag result in each time period after the time period; constructing a prognosis prediction model, and predicting the probability of occurrence of an endpoint event in each time period of the patient; constructing a loss function which comprises an error between a prediction result and a prognosis label result in each time period, and a prognosis evolution trend regularization ...

Подробнее
Номер записи: 69
30-05-2023 дата публикации

Fault prediction method and device based on multi-modal data, equipment and storage medium

Номер: CN116184988A
Автор: GUO YUANJUN, XIAO QINGE, TAN YONG, WU CHENGKE
Принадлежит:

The invention provides a fault prediction method and device based on multi-modal data, equipment and a storage medium, and relates to the field of fault prediction.The method comprises the steps that multiple pieces of abnormal operation data of target equipment are obtained, and the abnormal operation data correspond to different data types; performing data fusion on the abnormal operation data to obtain fused data; a prediction model is obtained, the fusion data are input into the prediction model, target fault prediction data are obtained, the prediction model comprises a plurality of sub-prediction models, the sub-prediction models are constructed based on different types of neural networks, and the weight value of each sub-prediction model is determined based on the data distribution characteristics of the fusion data. According to the method, the multi-modal data is adopted for prediction, and a plurality of different sub-prediction models are integrated in one prediction model, so ...

Подробнее
Номер записи: 70
01-08-2023 дата публикации

Cornea conus grading detection device

Номер: CN116525099A
Автор: CHEN XINJIAN, XU XIN
Принадлежит:

The invention relates to the technical field of medical artificial intelligence. The invention particularly relates to a cornea conus grading detection device. The device comprises an image base classifier for analyzing an acquired corneal topographic map, acquiring local features and global features of the required corneal topographic map, fusing the acquired local features and global features, and classifying fusion results. And the data base classifier is used for analyzing the obtained clinical features of the person needing to be detected and screening and classifying the clinical features. And the fusion module is used for performing fusion processing on classification results obtained by the image base classifier and the data base classifier to obtain a final classification result of the keratoconus of the person needing to be detected. According to the method, targeted modeling is performed on the corneal topographic map and the clinical features respectively, the problems of limitation ...

Подробнее
Номер записи: 71
04-08-2023 дата публикации

Time sequence classification method and device based on dictionary tree and coverage degree

Номер: CN116541784A
Автор: YANG JUN, JING SIYUAN
Принадлежит:

The invention relates to the technical field of data mining, in particular to a time sequence classification method and device based on a dictionary tree and coverage, and the method comprises the steps: obtaining a time sequence data set, namely a numerical time sequence data set; preprocessing the numerical time series data set to obtain a symbolized time series data set; a dictionary tree is constructed according to pattern features of the symbolized time sequence data set, and the pattern features are character sequences of all nodes on a path from a root node to any node; calculating an evaluation value of each node in the dictionary tree by adopting a coverage index; determining mode feature representatives based on the evaluation values of the nodes; obtaining a time sequence classifier based on the mode feature representation and the classifier model; and based on the time sequence classifier, classifying the to-be-classified time sequence data. According to the method, the time ...

Подробнее
Номер записи: 72
18-08-2023 дата публикации

Abnormal transaction recognition model construction method and abnormal transaction recognition method and device

Номер: CN116611507A
Автор: SUN TING, WU YANSHENG, LI JUAN, XIANG YONGXIANG
Принадлежит:

The invention relates to a construction method of an abnormal transaction recognition model, and an abnormal transaction recognition method and device, and is applied to the technical field of artificial intelligence, and the method comprises the steps: obtaining initial sample transaction data, carrying out the layering re-sampling of normal transaction data and abnormal transaction data in the initial sample transaction data, and obtaining target sample transaction data; obtaining a plurality of initial transaction features of the target sample transaction data, and screening out a plurality of target transaction features from the plurality of initial transaction features based on a first correlation coefficient between each initial transaction feature and transaction abnormality determination; constructing a plurality of decision trees according to the plurality of target transaction features to form a random forest model; and training each decision tree in the random forest model through ...

Подробнее
Номер записи: 73
03-10-2023 дата публикации

Learning data generating apparatus, learning data generating method, and non-transitory computer readable-storage medium

Номер: US0011775612B2
Автор: Shinji Yamamoto, Takato Kimura
Принадлежит: Canon Kabushiki Kaisha, CANON KABUSHIKI KAISHA

In order to provide a learning data generating apparatus that is able to efficiently restrain erroneous detections, the learning data generating apparatus includes a data acquisition unit configured to acquire learning data including teacher data, and a generation unit configured to generate generated learning data based on the learning data and a generating condition, wherein the generation unit converts teacher data of a positive instance into teacher data of a negative instance according to a preset rule when generating the generated learning data.

Подробнее
Номер записи: 74
25-08-2023 дата публикации

Method and device for dynamically recommending reserved network points, electronic equipment and medium

Номер: CN116644235A
Автор: LIU HAO, YANG CHEN, XU TAO, LIU CONGSHUO
Принадлежит:

The invention provides a method and a device for dynamically recommending a reservation network, electronic equipment and a medium, which can be applied to the technical field of big data and the technical field of artificial intelligence. The method comprises the following steps: acquiring time information when a customer enters a reservation system; based on the time information, judging the opening condition of a reservation entrance when the customer enters the reservation system; under the condition that the reservation entrance is not opened, outputting a first recommendation branch based on a pre-constructed beforehand recommendation model; and under the condition that the reservation entrance is opened, outputting a second recommended website based on a pre-constructed in-event dynamic adjustment model.

Подробнее
Номер записи: 75
18-08-2023 дата публикации

Multi-mode-based chronic obstructive pulmonary disease acute exacerbation prediction device

Номер: CN116612885A
Автор: YING HAOCHAO, DU BANG, XU HONGXIA, WU JIAN
Принадлежит:

The invention discloses a multi-modal-based chronic obstructive pulmonary disease acute exacerbation prediction device, and belongs to the technical field of computer-aided diagnosis. Structured diagnosis, examination and inspection data of the pneumology department and three-dimensional image data of a chest CT image are sent into a multi-base learning device module, a structured data classification module, a chest CT image segmentation module, a lung segmentation image feature extraction module, a fusion module and a prediction module. The optimized chronic obstructive pulmonary disease acute exacerbation prediction system is obtained through continuous optimization by adjusting parameters, and prediction is carried out according to feature information of modal fusion. The method is applied to a real medical scene, various complex examinations can be reduced, prediction can be carried out only by using personal information data, lung functions, blood routine and chest CT image data of ...

Подробнее
Номер записи: 76
29-08-2023 дата публикации

Hot pepper harvester roller rotating speed control method based on machine learning

Номер: CN116661514A
Автор: REGIN, ZHANG XINYU, QIN XINYAN, WANG BINGPENG, CHENG CHENMING, ZHAI ZHIYUAN, JIA WENXING
Принадлежит:

The embodiment of the invention provides a pepper harvester roller rotating speed control method based on machine learning, and the method comprises the steps: collecting a torque signal on a spring-tooth roller pepper harvester roller main shaft and the current roller rotating speed, inputting the collected torque signal into a torque prediction model, predicting the torque at the next moment, and calculating the rotating speed of the spring-tooth roller pepper harvester roller main shaft. Obtaining the load condition of the roller at the next moment through a mechanical model, adjusting the rotating speed of the roller until the predicted load condition of the roller is in a normal state, collecting a vibration signal of the spring tooth mounting plate, inputting the vibration signal into a pepper plant yield recognition model to recognize the current harvesting plant yield condition, and controlling the rotating speed of the roller according to a preset control method. According to the ...

Подробнее
Номер записи: 77
28-07-2023 дата публикации

Radar antenna scanning mode identification method and device, equipment and storage medium

Номер: CN116502152A
Автор: WANG RONG, WANG ZHENG, SHIN HO-NAM, DUAN YU, LU JITAO, NIE FEIPING, LI XUELONG
Принадлежит:

The invention provides a radar antenna scanning mode identification method and device, equipment and a readable storage medium. The identification method comprises the following steps: preprocessing acquired signal data to obtain a processed pulse amplitude data sequence; mining a preset characteristic pulse amplitude data sequence to judge whether a radar antenna scanning mode is a mechanical scanning mode or an electronic scanning mode; when the mode is the mechanical scanning mode, preset features of the preset feature pulse amplitude data sequence are extracted; and based on a pre-constructed classification model, identifying a specific scanning type in the preset feature mechanical scanning mode according to the preset features. The radar antenna scanning mode is easy to identify, the automation degree is high, the identification accuracy is high, the stability is good, and the real-time performance is good.

Подробнее
Номер записи: 78
29-08-2023 дата публикации

Data processing method and device, equipment and storage medium

Номер: CN116662909A
Автор: LU HUTAO, JIA DUAN, TIAN KUO, YANG XIA
Принадлежит:

The invention provides a data processing method and device, equipment and a storage medium, and can be applied to the technical field of big data. The method comprises the steps that the transaction volume and multiple pieces of first market information of a target object in a historical period are obtained, the storage structure of the multiple pieces of first market information is a linear storage structure, a target dynamic balance tree type is determined from multiple dynamic balance tree types according to the transaction volume, and the target dynamic balance tree type is determined according to the target dynamic balance tree type. A target balance tree corresponding to the multiple pieces of first market information is generated, the target balance tree is used for storing the multiple pieces of first market information, the storage structure of the target balance tree is a tree-shaped storage structure, the multiple pieces of first market information are sorted according to the ...

Подробнее
Номер записи: 79
21-04-2023 дата публикации

Public power resource allocation method based on big data

Номер: CN115995814A
Автор: ZHU SHAOHUA, HWANG DONG-UK, PAN LIQUN, LI YOUYI, ZOU ZHENDONG
Принадлежит:

The invention provides a public power resource allocation method based on big data, and belongs to the technical field of power resource allocation, and the method comprises the steps: predicting the standard power consumption of each resident in a subsequent stage; matching the average big data of the electricity consumption of the residents in the regulation and control area, and correcting the standard electricity consumption of each resident; and allocating power resources according to the corrected standard power consumption. According to the invention, the problem that the distribution scheduling is unreasonable because the electric quantity is distributed to some inactive users due to the fact that the current electric power distribution mode does not further realize the accurate prediction of the power consumption of the users in the subsequent stage is effectively solved.

Подробнее
Номер записи: 80
01-08-2023 дата публикации

Lithology identification method and system applied to oil drilling

Номер: CN116522251A
Автор: TANG HUAFENG, BIAN WEIHUA, GAO YOUFENG, ZHANG YAN, HUANG YULONG, WANG PUJUN
Принадлежит:

The invention discloses a lithology identification method and system applied to oil drilling, and the method comprises the steps: collecting geophysical data in a construction site before the drilling process; preprocessing the geophysical data to obtain processed data; performing feature extraction on the processed data to obtain a training set and a test set; constructing a lithology classification model by using the training set and the test set; using the lithology classification model to complete lithology prediction before the oil drilling process to obtain a prediction result; drilling based on the prediction result; in the drilling process, rock ruler data and rock debris data are collected; and lithology identification is completed based on the rock ruler data and the rock debris data. According to the method, various geophysical data are adopted, data processing and an artificial intelligence algorithm are comprehensively applied for analysis and processing, so that the geological ...

Подробнее
Номер записи: 81
15-08-2023 дата публикации

Emergency traffic transportation guarantee system, method and equipment based on geographic data

Номер: CN116596423A
Автор: ZHANG FASHAN, WANG GANG, PAN BIN, ZHANG SHUAI, XU KAIKAI, LUO HUAISHEN, ZHANG ZHENDONG, BAI JIE
Принадлежит:

The invention discloses an emergency traffic transportation guarantee system, method and equipment based on geographic data, mainly relates to the technical field of traffic transportation guarantee, and is used for solving the problem that personnel cannot be quickly dispatched by an existing method. Comprising a data receiving module used for acquiring geographic reference data and event processing data and binding event uploading data with the geographic reference data; the data classification module is used for performing parent classification on the geographic parameter data according to the similarity of the route map; according to an event occurrence environment and an event level of event processing data corresponding to the geographic parameter data, performing sub-classification on the geographic parameter data of the same parent classification; and the emergency notification module is used for acquiring the emergency report information, determining the corresponding geographic ...

Подробнее
Номер записи: 82
01-08-2023 дата публикации

Method and device for carrying out crowd packet statistics by using federated learning model

Номер: CN116521952A
Автор: YOU ZHIQIANG, BIAN YANG, WANG ZHAOKAI, ZHANG WEIQI
Принадлежит:

The embodiment of the invention provides a method and device for carrying out crowd packet statistics by using a federated learning model. The federal learning model comprises a plurality of trees. A first participant participating in federated learning has a weight matrix of leaf nodes of each tree in a plurality of trees. The second participant participating in federated learning has a prediction result matrix generated by each tree in the multiple trees for the crowd packet. The method is performed by a first participant. The method comprises the following steps: splicing weight matrixes of a plurality of trees into a first splicing matrix; obtaining a second splicing matrix generated by a second participant, wherein the second splicing matrix is generated by splicing the prediction result matrixes of the multiple trees according to columns and executing an out-of-order operation according to the columns; performing matrix multiplication on the first splicing matrix and the second splicing ...

Подробнее
Номер записи: 83
25-07-2023 дата публикации

Probabilistic earthquake and tsunami disaster analysis method based on logic tree method

Номер: CN116482763A
Автор: BAI YEFEI, LIU JINWEI, ZHI HONGHUAN, ZHOU SANGJUN, ZHOU YIFAN, ZHAO WENYU, WEI XIAORAN
Принадлежит:

The invention relates to a probabilistic earthquake and tsunami disaster analysis method based on a logic tree method, and the method comprises the steps: obtaining a to-be-analyzed region, and enabling the to-be-analyzed region to be used for the probabilistic earthquake and tsunami disaster analysis method; according to the to-be-analyzed area, respectively determining an earthquake tsunami source and a high-precision topographic file; constructing a parameter logic tree according to the first parameter group and the second parameter group corresponding to the earthquake tsunami source; calculating by using a tsunami numerical model according to the high-precision terrain file, and determining a tsunami maximum wave height file; and according to a logic tree method formed by the parameter logic tree and probabilistic combination and calculation of the plurality of maximum wave height files, determining a wave height exceeding probability at each pixel point of the to-be-analyzed area.

Подробнее
Номер записи: 84
14-07-2023 дата публикации

Village public data intelligent management system

Номер: CN116431815A
Автор: ZHAO BIN, ZHANG MIN, ZHANG WENYIN, GAO YILONG, WANG LEI
Принадлежит:

The invention relates to the technical field of digital data processing, in particular to a village public data intelligent management system. The system comprises a village data public platform database, an entry capturing module, a data processing module and an information management module, wherein the data processing module further comprises a keyword labeling module, a keyword decision index module, a public information vector processing module, a phrase induction module, a phrase weight module and a decision optimization ratio module. According to the method, the situation that words with the same frequency in the disclosed village information reduce the classification accuracy of the whole village information can be avoided, and the follow-up measurement of the similarity degree between the village information is facilitated. And secondly, a decision optimization ratio is constructed based on public vectors, so that the influence of a small amount of words in special village public ...

Подробнее
Номер записи: 85
28-07-2023 дата публикации

Broadband oscillation suppression method, medium and system for shore-based power supply system

Номер: CN116505552A
Автор: ZOU ZHIXIANG, TANG JIAN, YANG ZHONGGUANG, DOU ZHENLAN, WANG ZHENG, CHENG MING
Принадлежит:

The invention discloses a broadband oscillation suppression method, medium and system for a shore-based power supply system in the technical field of broadband oscillation suppression, and the method comprises the steps: obtaining the frequency of broadband oscillation generated by the shore-based power supply system and corresponding characteristic data, inputting the data into a classifier model as a training set, generating a decision tree classification sub-model by the classifier model; inputting the actually obtained frequency and amplitude of broadband oscillation generated by the shore-based power supply system and the equivalent external characteristics of the power supply system into the classifier model after the decision tree classification sub-model is generated, so that the classifier model outputs a broadband oscillation suppression strategy based on classification conditions; and inputting the broadband oscillation suppression strategy into the first neural network, outputting ...

Подробнее
Номер записи: 86
25-05-2023 дата публикации

MACHINE LEARNING PIPELINE OPTIMIZATION

Номер: US20230162289A1
Автор: Alain Charles Briancon, Jean Joseph Belanger, Chris Coovey, Valisis Sotiris, Eric Simon
Принадлежит:

Provided is a process of modeling methods organized in racks of a machine learning pipeline to facilitate optimization of performance using modelling methods for implementation of machine learning design in an object-oriented modeling (OOM) framework, the process including: writing classes using object-oriented modelling of optimization methods, modelling methods, and modelling racks; writing parameters and hyper-parameters of the modeling methods as attributes as the modeling methods; scanning modelling racks classes to determine first class definition information; selecting a collection of rack and selecting modeling method objects; scanning modelling method classes to determine second class definition information; assigning racks and locations within the racks to modeling method objects; and invoking the class definition information to produce object manipulation functions that allow access the methods and attributes of at least some of the modeling method objects, the manipulation functions ...

Подробнее
Номер записи: 87
17-10-2023 дата публикации

Method of building a dynamic overlay network topology based on cross-chain interaction between blockchains and device therefor

Номер: US0011792105B2
Автор: Jiang Xiao, Shijie Zhang, Jiajie Zeng, Hai Jin, Chen Yu
Принадлежит: HUAZHONG UNIVERSITY OF SCIENCE AND TECHNOLOGY, Huazhong University of Science and Technology

The present invention relates to a method of building a dynamic overlay network topology based on cross-chain interaction between blockchains, at least comprising: selecting at least one gateway node and at least one ordinary node from at least one cross-chain interaction node; clustering the ordinary nodes into ordinary node clusters based on a clustering algorithm; building a structured network topology and cross-chain connections between the cross-chain interaction nodes based on the gateway node; performing topology building on the ordinary node clusters based on a breadth-first spanning tree algorithm.

Подробнее
Номер записи: 88
25-07-2023 дата публикации

Parameter processing method and device of display screen module, equipment and storage medium

Номер: CN116484269A
Автор: NI ZHENGHUA
Принадлежит:

The invention relates to the field of artificial intelligence, and discloses a display screen module parameter processing method, device and equipment and a storage medium, which are used for realizing intelligent display screen module parameter analysis and improving the solving accuracy of optimal operation parameters of a display screen. The method comprises the steps of performing feature correlation analysis on multiple pieces of feature parameter data to obtain at least one parameter correlation evaluation index corresponding to every two pieces of feature parameter data; constructing a target parameter feature analysis model corresponding to every two pieces of feature parameter data according to the at least one parameter correlation evaluation index, and obtaining standard parameter data corresponding to each piece of feature parameter data; respectively inputting the standard parameter data and the plurality of characteristic parameter data into a target parameter characteristic ...

Подробнее
Номер записи: 89
04-08-2023 дата публикации

Bidirectional single-lane intelligent meeting scheduling method in adit based on machine learning

Номер: CN116543556A
Автор: FU LIANGRUI, LI ZENGYAN, CHEN TAO, DENG JINQIU, DAI JIAKAI, ZHU BAOLIANG, PAN RONGMIN
Принадлежит:

The invention discloses a two-way single-lane intelligent meeting scheduling method in an adit based on machine learning, and relates to the field of mine excavation engineering management and vehicle scheduling, and the method comprises the following steps: 1, collecting and preprocessing key information of vehicles in the adit through a meeting situation awareness thread, and caching the information; 2, further polling vehicle cache information in real time by a vehicle meeting early warning thread; and step 3, calling a preset comprehensive decision model, and automatically issuing a meeting scheduling instruction in time or prompting manual intervention of meeting scheduling, according to the method, different types of learning models (empirical functions, supervised learning and unsupervised learning) are fused and complement each other, the empirical functions can represent general rules, and the method is efficient and concise, and is suitable for large-scale popularization and application ...

Подробнее
Номер записи: 90
11-08-2023 дата публикации

Random forest-based link quality evaluation method for cooperative network with and without people

Номер: CN116578932A
Автор: XIAO ZHE, WANG YANGYANG, JIAO LIBIN, LIU LIZHE, GAN RUIMENG, JIA ZEKUN, LIU YONGJIANG
Принадлежит:

The invention provides a random forest-based link quality evaluation method for an unmanned cooperative network, and belongs to the field of wireless communication networks and unmanned cooperative ad hoc networks. The method comprises the following steps: firstly, constructing a link quality evaluation index system; then constructing a random forest model, and balancing sample data through a stratified sampling method; optimizing hyper-parameters of the random forest model by adopting a sparrow search algorithm, improving a discoverer position updating method in the sparrow search algorithm by introducing a butterfly optimization algorithm, and limiting a search space of a population in each iteration by adopting a self-adaptive reduction search space mechanism; and finally, realizing network link quality evaluation based on the random forest classification model optimized by the improved sparrow search algorithm. According to the method, the accuracy and the convergence speed of link ...

Подробнее
Номер записи: 91
08-08-2023 дата публикации

Method and terminal for detecting vehicle entering service area based on ETC (Electronic Toll Collection) data

Номер: CN116561680A
Автор: ZOU FUMIN, HUANG SHIBIN, GUO FENG, CAI QIQIN, HUANG CHUNBIN, YU XIANG, XU GEN, WU SONGYANG, LIN ZIYANG, REN QIANG, WANG HAOLIN, ZHUANG YUANZHENG, XIE MINGWEI, LIN YAYI
Принадлежит:

The invention discloses a method for detecting vehicle entering a service area based on ETC data. A driving track of a vehicle in a road section where a certain service area is located is acquired according to ETC transaction data; acquiring data of a vehicle driving into the service area, and constructing a first data set according to the driving track by taking whether the vehicle enters the service area as a feature; respectively calculating the driving speeds of the vehicle in the front section, the service area section and the rear section of the service area, and constructing a second data set by taking the driving speed of each section of the vehicle as a feature; and establishing a classification model according to the first data set and the second data set, training, and judging whether the vehicle enters the service area or not by using the classification model after the training is completed. Whether the vehicle enters the service area or not is judged according to the ETC transaction ...

Подробнее
Номер записи: 92
06-06-2023 дата публикации

Advanced data collection block identification

Номер: US0011669588B2
Автор: Martynas Juravicius, Andrius Kuksta
Принадлежит: Oxylabs, UAB

Systems and methods that allow examination of response data collected from content providers and provide for classification and routing according to the classification. The process of classification employs an unsupervised, or partially unsupervised, Machine Learning classifier model for identifying data collection responses that contains no data, mangled data, or a block, for assigning a classification correspondingly and for feeding the classification decision back to a data collection platform.

Подробнее
Номер записи: 93
07-11-2023 дата публикации

Method of threat detection in a computer network security system

Номер: US0011811788B2
Автор: Matti Aksela
Принадлежит: WITHSECURE CORPORATION, F-Secure Corporation

A method comprising: receiving raw data related to one or more network nodes, wherein dissimilar data types are aligned as input events; filtering one or more of the input events by using an adjustable threshold that is based on a filtering score, wherein the filtering score is an estimate of the likelihood that the input event is followed by a security related detection; processing only input events passed through the filtering by an enrichment process; and analysing the data received from the enrichment process for generating a security related decision.

Подробнее
Номер записи: 94
27-06-2023 дата публикации

Characteristic selection method based on random forest improvement

Номер: CN116340746A
Автор: ZHOU WENJIN, AN YUNFEI, MIAO SHIDI
Принадлежит:

The invention discloses a feature selection method based on random forest improvement. The feature selection method comprises the following steps of firstly performing missing value processing and discretization processing on a complete data set; secondly, establishing a traditional random forest model, calculating the accuracy rate of the traditional model, then calculating the optimal tree depth when the number of trees is set, regenerating the random forest with the optimal depth, calculating the accuracy rate of each tree in the newly generated forest, selecting a certain percentage of trees with the higher accuracy rate, and then calculating the data similarity of all the trees to obtain the random forest. Removing the trees with the similarity exceeding a set value and small accuracy, and finally calculating the final accuracy so as to obtain an improved random forest. The reason for performance reduction of the classifier is often that the high-dimensional features contain irrelevant ...

Подробнее
Номер записи: 95
15-08-2023 дата публикации

Evaluation method and application of clean energy supply system based on big data

Номер: CN116595460A
Автор: YUAN LING, ZHANG JIXIN, YANG JINGANG, SUN ZHIFENG
Принадлежит:

The invention belongs to the technical field, and discloses a big data-based clean energy supply system evaluation method and application, and the system comprises a performance data collection unit which is used for collecting the performance data of the clean energy supply system during the operation period; a data transmission unit; the classification unit is used for determining the type of environmental influence generated by the clean energy supply system during operation; the evaluation unit is used for evaluating the influence of the clean energy supply system on the environment; the energy supply unit is used for determining energy supply information according to an evaluation result of the evaluation unit by using an energy supply evaluation program; and the data storage unit is used for storing the data. According to the method, various types of environmental influences of different clean energy systems can be quantitatively, objectively and accurately analyzed, the main environmental ...

Подробнее
Номер записи: 96
14-07-2023 дата публикации

Classification method and device, electronic equipment and storage medium

Номер: CN116432124A
Автор: POLVALI MOHSEN, SHENG CHEN
Принадлежит:

The embodiment of the invention discloses a classification method and device, electronic equipment and a storage medium, and the method comprises the steps: carrying out the feature extraction of a target object, obtaining the target feature of the target object, carrying out the processing of the target feature, and obtaining a class recognition result of the target object; wherein the category recognition result of the target object is matched with a first target logic rule determined based on a classification tree; the first target logic rule represents a logic relationship between the category of the target object in the classification tree and the ancestor category of the category of the target object.

Подробнее
Номер записи: 97
08-08-2023 дата публикации

Fusion type fault diagnosis method based on digital twinning

Номер: CN116561681A
Автор: TANG JING, JI HONGBING, ZHANG WENBO, LI LIN, ZANG BO, WANG HAIJIAN, XU ZENGZHI
Принадлежит:

The invention discloses a fusion type fault diagnosis method based on digital twinning. The method comprises the following steps: (1) establishing an equipment digital twinborn model; (2) updating the digital twin model of the equipment; (3) operating the digital twin model of the equipment, and simulating the fault operation state of the equipment; (4) establishing a random forest fault diagnosis model, and outputting equipment fault classification and fault prediction results; (5) preparing a verification data set, and recording and evaluating the performance and indexes of the random forest fault diagnosis model; (6) evaluating model performance indexes; (7) adopting a fusion type fault diagnosis algorithm, using the optimized fault diagnosis model to correct the output of the digital twinborn model, and evaluating the performance index of the model; (8) judging whether the unmanned aerial vehicle has a fault or not according to an output result; according to the invention, the purposes ...

Подробнее
Номер записи: 98
28-07-2023 дата публикации

Eddy core parameter calculation method and system based on Doppler radar detection

Номер: CN116500580A
Автор: GU RUNPING, LU TONG, WEI ZHIQIANG
Принадлежит:

The invention relates to the technical field of civil aviation air traffic control, and provides a vortex core parameter calculation method and system based on Doppler radar detection. The method comprises the following steps: acquiring detection data of a tail vortex field through a radar and eliminating background wind influence; according to the pure data, characteristic parameters, including the standard deviation of the radial speed and the elevation angle, of the distance gate are obtained through calculation; establishing a decision tree based on velocity field characteristics near the vortex core, and inputting the characteristic parameters into the decision tree to obtain a radial distance range of the vortex core; optimizing the vortex core coordinate and the circulation by taking the minimum mean square error as an objective function to form a pre-vortex core parameter; and inverting the pre-vortex core parameter based on the radar detection time and the flight time to obtain ...

Подробнее
Номер записи: 99
07-07-2023 дата публикации

Prediction model generation method and system based on commodity data

Номер: CN116402544A
Автор: GAO BO, BAI JINGTAO
Принадлежит:

The embodiment of the invention provides a prediction model generation method and system based on commodity data, and the method can obtain historical data including transaction time, the number of commodities, unit price and total transaction amount, and generates time expansion data according to the transaction time. Wherein the time expansion data comprises a transaction time interval, a median of the transaction time interval and an average of the transaction time interval. And segmenting the historical data and the time expansion data into a training set and a verification set, and training an initial prediction model generated based on an XGBoost algorithm through the training set to obtain a data prediction model. And inputting the verification set into the data prediction model to obtain a verification result of the data prediction model. If the verification result is over-fitting or under-fitting, the hyper-parameter and the feature weight of the data model are adjusted, and the ...

Подробнее
Номер записи: 100