13-06-2023 дата публикации
Номер: CN116257738A
Принадлежит:
The embodiment of the invention provides a brain load classification prediction method and system based on package-type dimensionality reduction, and the method comprises the steps: carrying out the preprocessing of collected electroencephalogram signals for training set data, extracting the average power spectrum density under the four rhythms of delta, theta, alpha and beta, reducing the feature number to an intermediate dimension, outputting a dimensionality reduction matrix, and building a brain load classification model; and for the test set data, performing the same processing on the electroencephalogram signals of the test set according to the training set data, and predicting the mental load condition of the operator in the occurrence period of the electroencephalogram signals of the test set by using the mental load classification model. According to the technical scheme, the data dimension can be reduced, and on the premise that dimensionality disasters are relieved to a certain ...
Подробнее