In the realm of finance, accurately predicting stock market trends has always been a formidable challenge. However, with the emergence of machine learning as a powerful tool for forecasting, this research paper undertakes a comparative analysis of four machine learning models in terms of their accuracy in predicting the short-term performance of three popular stocks traded on the NYSE between March 2020 and May 2022. The models employed and refined in this study include XGBoost, Random Forest, Multi-layer Perceptron, and Support Vector Regression. Our evaluation metrics, namely RMSE, MAPE, MTT, and MPE, help identify the models that exhibit the highest level of accuracy. By utilizing a training data set spanning 240 trading days, we discover that XGBoost yields the greatest accuracy, despite its longer execution time of up to 10 seconds. It is worth noting that further refinement of individual parameters or the inclusion of additional exogenous variables may enhance the results of this study.
Using exogenous variables and machine learning algorithms to forecast short-term stock prices.
