Predicting the Direction of Stock Price Movement with Machine Learning Algorithms

Abstract

The application of machine learning algorithms in predicting stock price directional movements has been a widely discussed problem. Due to the chaotic, uncertain, and dynamic characteristics of stock markets, relatively accurate predictions could contribute to financial benefits and risk reduction. This paper aims to assess the prediction performance of five well-known machine learning models, which are Logistic Regression, Support Vector Machines, Decision Trees, Random Forest, and Extreme Gradient Boosting. Data in this study covers the period from 2020-01-02 to 2023-07-03, and assets studied in this paper are Nike, Amazon, Microsoft, Tesco, and Airbnb. After feature selection, data pre-processing, cross-validation, model selection, and evaluation, it is concluded that the Random Forest Classifier tends to perform better in directional predictions, as it demonstrates higher accuracy and precision. This research highlights the application of machine learning algorithms in financial area, especially in the stock price prediction.