Stock-Price-prediction-using-Recurrent-Neural-Network-LSTM-

In this project, I made an attempt to build a LSTM-RNN model to predict stock prices using keras with tensorflow(backend). The training data comes from historical closing prices of various stock indices and news sentiment score. The accuracy of the stock price prediction is measured by Root Mean Square Error (RMSE). We did some experiments on the network's hyper-parameters such as LSTM cell hidden state size, truncated back propagation length and depth of the network. Last but not the least, we built a website using this prediction model as engine with Flask and python.