The purpose of this project is to measure the effect that sentiment has on stock prediction software. This project utilizes the Keras library in Python to build LSTM models. After the LSTM models are trained and tested, one LSTM model will be given a sentiment feature, and the accuracy of each LSTM to true market values will be evaluated.
Refactored get_data so that instead of data being used then thrown away, it is stored in a database to reduce redundancy.