Using Machine Learning To Predict Future Stock Prices

Abstract

The volatile nature of financial markets presents a formidable challenge in accurately forecasting equity returns and stock prices. Artificial intelligence (AI) and machine learning are increasingly being employed to predict future stock prices and reduce risks, enabling more profitable investment decisions.  

In this paper, we employed a few AI regression and classification models to tackle the challenge of predicting future stock prices These models utilized a historical time-series dataset of the opening and closing prices of stock over the last five years to generate predictions that were eventually compared with actual prices The models were evaluated based on their regression or classification score, measured as a percent of accuracy. Among the models tested the MLP classification and linear  regression models were the most effective predictors.