GitHub Backtesting - Linear Regression & Classification algorithm applied to GLD & AMZN

Nota : Please click on the Github image above in order to be redirected to the project.

Overview

For this case study, we backtested linear regression and classification algorithm (logistic) to respectively SPY and AMZN, from January 2015 to May 2022. The results showed that logistic regression, standard classification algorithm, has better results in the prediction of the direction of market movements than linear model with and without transaction costs.