Comparison of forecasting methods. Within this jupyter notebook I perform some basic long term stocks forecasting by running several different ML frameworks and comparing them to an LSTM approach as well as the classic ARIMA method. This project showcases how ARIMA still remains one of the most useful approaches for basic forecasting.
The work could be further improved by incorporating trend and momentum indicators into the models. However, I decided not to include them in the end to leave room for comparison between these current models and the modified ones.