Algorithm 1: EEG Fatigue Recognition Engine

Input: EES Output: classification accuracy Acc 1: Begin 2: import EES to python 3: Read the EES and select the signal of two of the channels

4: X $\leftarrow$ EES; H $\leftarrow$ 0.15 Hz; L $\leftarrow$ 40 Hz; D $\leftarrow$ 50 Hz; 5: for each of EES do

6:  PEES = bandpass filtering(X, H, L);

7:  PEES = depressionfiltering(PEES, D);

8: end for

9: for each of PPS do

10:  FF = calculate frequency (fourier transform(PEES));

11:  EF = calculate entropy (PPS); 12: end for 13: train_label, train_data = Select 80% of data(FF + EF); 14: test_label, test_data = Select 20% of data(FF + EF);

14: model = LightGBMtrain (train_label, train_data);

15: classification accuracy as Acc = LightGBMdict(test_label, test_data, model);