摘要：Depression is a kind of relatively common psychological disease of among people. The extractof EEG feature is to utilize the course of development of better aided diagnosis with depression patients, Soas to put forward the accurate treatment options. The traditional machine study is to directly input EEG intoNeural Networks and not to consider the influence of time series for data accuracy and Bi-LSTM is not onlyto inherit the treatment of LSTM to timely constraint, but also combine the influence of twoway factors onneutral netwotk, which has good computing advantage. This essay adopts a kind of the study of EEGrecognition of depression on Bi-LSTM based on ERP. Compared with other model, the accuracy rateidentification and classification under 16 reaches 80.6% with good credit after the improvement of the Bi-LSTM.
其他摘要：Depression is a kind of relatively common psychological disease of among people. The extract of EEG feature is to utilize the course of development of better aided diagnosis with depression patients, so as to put forward the accurate treatment options. The traditional machine study is to directly input EEG into Neural Networks and not to consider the influence of time series for data accuracy and Bi-LSTM is not only to inherit the treatment of LSTM to timely constraint, but also combine the influence of two-way factors on neutral network, which has good computing advantage. This essay adopts a kind of the study of EEG recognition of depression on Bi-LSTM based on ERP. Compared with other model, the accuracy rate identification and classification under 16 reaches 80.6% with good credit after the improvement of the Bi- LSTM.
