The purpose of this project is to take on the Winton Stock Market Challenge on Kaggle sponsored by Winton Capital. The aim of this challenge is to predict intra and end of the day returns using already given historical stock performance and a host of masked features.
The different approaches adopted in this project utilize regression using Neural Networks to make future predictions of stock returns, whose performance is measured by loss and Mean Square Error. In order to predict the future stock returns, given the data, the best fitting models would be Long Short Term Memory Networks and the Sequential Model.