Desarrollo de un modelo de predicción para los valores de las acciones del índice S&P500, sobre la base del estado del arte de la temática, usando un enfoque de análisis de datos, que incluya técnicas de minería de datos y de aprendizaje automático.

Abstract

The work proposes an approach based on LSTM (Long Short-Term Memory) neural networks to predict the actions of the S&P500, one of the most important indexes of the stock market. Predicting prices in the stock market is a complex challenge due to its dynamic and volatile nature. The model is trained using a historical data set of the S&P500 from January 2015 to May 2023, which includes information on open, close, high, and low prices, among others. The experimental results show that the proposed model outperforms other traditional prediction approaches. The LSTM-based approach achieves higher accuracy in S&P500 price prediction, which can be of great use to investors and financial analysts in making investment decisions. However, it is suggested that further research be done to explore other variants of LSTM models and further improve the prediction accuracy.