Predicting S&P500 index using LSTM networks

Files/Directories included in the main directory:

• research/LSTM_Prediction.ipynb - Jupyter notebook containing instructions for training the LSTM network and forecasting the prices.

• src/prediction.py - Python script for forecasting prices

• src/train_lstm_model.py - Python script for training/updating the LSTM model

Instructions:

• Install all the dependencies from requirements.txt

• For updating or training the LSTM model with newer prices, use a csv file containing Closing prices

• For training/updating use the command, 'python train_lstm_model.py' and type the name of csv file (or path of csv file)

• For forecasting use the command, 'python prediction.py' and enter the number of days you want to forecast (It is recommended to enter a value less than 15 days and always traing the network on a 15 days interval)