Unbalanced electricity stealing data classification method and device based on SMOTE-GBDT, computer equipment and storage medium

The invention discloses an SMOTE-GBDT-based unbalanced electricity larceny data classification method and device, computer equipment and a storage medium. The method comprises the steps of 1, collecting user load data; 2, carrying out missing value filling and standardization processing on the data; step 3, taking different K oversampling based on an SMOTE algorithm; 4, training a GBDT model by adopting GBDT of default parameters for the oversampled data set, and 5, evaluating the classification performance of different K value generation data sets, and finding out the K neighbor value of optimal classification; step 6, training a GBDT model based on the data set with the best oversampling effect; and step 7, based on network search and cross validation combination model parameters, testing to find out optimal parameters of the GBDT model, and further obtaining an optimal electricity stealing analysis model. The electricity stealing behavior analysis of the user can be realized through ...