Stock-Prediction-Linear-Regression

Jupyter Notebook that uses 3 different regression methods to predict closing stock prices for each day. Features used include previous H/L/O/C, competitor stock prices and rolling correlation, moving averages, and return. Future ideas for this project include the incorporation of better technical indicators, standardising predictions to RETURN instead of absolute closing price, more advanced prediction mechanics (like weighting the coefficients of competitors' share price predictors by rolling correlation), sentiment analysis of the stock on twitter, and the building of a Flask web app.