Algorithm 2: ECG Fatigue Recognition Engine

Input: ECS Output: classification accuracy Acc 1: Begin 2: import ECS to python

3: Y $\leftarrow$ ECS; Hi $\leftarrow$ 0.1 Hz; Lo $\leftarrow$ 30 Hz; T $\leftarrow$ 0.3; 4: PECS = bandpass filtering (Y, Hi, Lo); 5: PECS = Trend processing (PECS); 6: QRS_Filter = sin (1.5 $\ast$ $\pi$, 3.5 $\ast \pi$ 15); 7: for each of PECS do 8:  SIM = correlate (PECS, QRS_Filter) 9:  if SIM > T 10:   RI = index (PECS); 11:   HF = calculate HRV feature (RI); 12:  end if 13: end for 14: train_label, train_data, test_label, test_data = Select 80% of data(HF); 15: test_label, test_data = Select 20% of data(HF); 16: model = LightGBM train (train_label, train_data); 17: classification accuracy as Acc = LightGBM dict (test_label, test_data, model);