Implemented by reference papers Stock Market Prediction Based on Generative Adversarial Network https://www.sciencedirect.com/science/article/pii/S1877050919302789
Based on this, we try to use CNN in the D network part.
Generator: LSTM model
Discriminator: MLP or CNN model
#parameters can be default or set by yourself.
python train.py -e[epochs] -m[D net model(mlp or cnn)] -t[timewindow] -d[dataset] -l1[hyper-parameters λ1] -l2[hyper-parameters λ1] -b[batch size] -lr[learning rate] -be[beta1] -s[split dataset rate]