Enhancing Holt-Winters Forecasting of PSEi Data With Genetic Algorithm and Cuckoo Search Algorithm: A Comparative Analysis

Accurate forecasting of stock market indices plays a crucial role in investment decision-making. The study investigates the application of the Holt-Winters forecasting method using two optimization algorithms, namely Genetic Algorithm (GA) and Cuckoo Search Algorithm (CSA), to enhance the forecasting performance of the Philippine Stock Exchange Index (PSEi). The Holt-Winters method is a popular technique for time series forecasting, capable of capturing level, trend, and seasonality components. GA and CSA are metaheuristic optimization techniques that can improve the accuracy of the Holt-Winters forecasts by optimizing the model parameters. The optimization models are assessed using historical PSEi data spanning several years. The experimental results show that GA outperformed CSA in the goal of increasing the Holt-Winters forecasting method's accuracy. The comparative analysis shows that both algorithms have promising results with little divergence in the outcomes, though there are differences in how well they can optimize the model parameters and recognize the intricate patterns in the PSEi data. The results imply that employing GA in the Holt-Winters forecasting model can be a useful strategy for stock market forecasting, empowering investors to make more informed choices.