This project is meant to test two ways to potentially choose stocks using probability analysis on the returns and a RNN (recurrent neural network) to predict future stock prices. The specific RNN used was a LSTM (long short-term memory). First 20 stocks with a similar P/E ratio, a metric used to predict future growth, are selected using data from yahoo finance. Of this 20, 5 stocks each were picked using the LSTM and stock return predictions.