Forecasting of Philippine Stock Exchange Index Using Optimized Artificial Neural Networks with Enhanced PSO Algorithm

Abstract

This research paper explores the application of the Enhanced Particle Swarm Optimization (PSO) algorithm, called the Random Adaptive Backtracking Particle Swarm Optimization (RAB-PSO) algorithm, to optimize Artificial Neural Networks (ANNs) for forecasting the Philippine Stock Exchange Index (PSEi). The study utilizes a dataset spanning from May 11, 2018, to May 10, 2023, sourced from Yahoo Finance and standardized for analysis. The hyperparameter of an ANN model is fine-tuned using the RAB-PSO algorithm to enhance forecasting accuracy. Evaluation metrics such as Root Mean Square Error (RMSE) and R2 are employed to assess the performance of the optimized ANN model. The results indicate that the ANN model optimized with RAB-PSO has minimal error rates, significantly outperforming the standard PSO algorithm. Generally, this research contributes to the field of PSEi forecasting and emphasizes the significance of optimizing hyperparameters through enhanced PSO for ANNs in financial prediction tasks.