Optimizing Stock Market Predictions Using Machine Learning Algorithms

Soto A.; Ticona W. · Lecture Notes in Networks and Systems (2025) · DOI: 10.1007/978-3-031-96759-7_31

Resumen

Stock prices are highly volatile due to various fluctuations, making them difficult to predict, especially for new investors. This research enhances stock price prediction accuracy by using innovative multimodal models that integrate multiple machine learning algorithms. The study followed four phases: dataset extraction, preprocessing, modeling with techniques such as LSTM, XGBoost, RF, KNN, DT, and SVM, and evaluation using metrics like RMSE, MAE, MAPE, MSE, and R2. The SVM model achieved the highest accuracy for companies such as Alicorp (99.88%), Credicorp (99.79%), Intercorp (99.70%), Minera Buenaventura (99.85%), and Minera Cerro Verde (99.89%). Additionally, Alicorp demonstrated excellent performance with MAE, MAPE, MSE, and RMSE values of 0.022, 0.333%, 0.001, and 0.042, respectively, showcasing its superior predictive capacity and minimal errors. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

Instituciones: Universidad Tecnológica del Perú