tanvird3/Dhaka_Stock_Exchange_Forecasting

Forecasting stock prices with Long Short Term Memory (LSTM) architecture

Forecasting Instruments of Dhaka Stock Exchange using Long Short-Term Memory Networks

Code Explanation

1. Codes inside LSTM_DSE_FORECAST.ipynb work with tensorflow 1.14.0
2. Codes inside LSTM_DSE_FORECAST_tensorflow2x.ipynb work with tensorflow 2.1.0 (Check requirements.txt for module list and versions)
3. Codes inside LSTM_DSE_FORECAST_tensorflow_latest.ipynb work with tensorflow 2.5.0 (for the time being)
4. You can launch LSTM_DSE_FORECAST_tensorflow_latest.ipynb on Google Colaboratory with the link provided above

Disclaimer

LSTM is a memory hungry procedure. Set TRIAL and PATIENCE at modest level for good predictions and be patient because it is going to take a while.