I am working on LSTM RNN on stock prediction sample data. It seems the RNN is not giving more weights to recent data. (This normally happens in ARIMA). The weights is equall