Title: Predicting future stock market trends using a hybrid CBR model and Aquila optimiser

Authors: Ying Song

Addresses: Economics & Management Department, Liuzhou Institute of Technology, Liuzhou, Guangxi, China

Abstract: The stock market forms a part of a country's economy and works as a mechanism for trading in publicly traded companies' shares. Financial gain maximisation is a function of changing factors, and therefore, a grasp of trends in the market is imperative. With a growing number of investors involved in trading, the skill for predicting stock price behaviour gained prominence for investors. Stock price prediction is a challenging problem in view of its complex, non-linear and uncertain nature in the marketplace. A new hybrid model utilising categorical boosting regression and optimisers for predicting stock price is proposed. Historical data for 2018 to early 2024 for the Korea Composite Stock Price Index (KOSPI) was analysed in detail. Performance evaluation for model efficacy included factors such as MAE, MAPE, RMSE and R². AO-CBR model produced an R²-value of 0.9952, representing high performance with minimum error compared with competing models.

Keywords: stock prediction; categorical boosting regression; KOSPI market; Aquila optimiser; financial market.

DOI: 10.1504/IJIPT.2024.146273

International Journal of Internet Protocol Technology, 2024 Vol.17 No.4, pp.180 - 190

Received: 17 Feb 2025
Accepted: 23 Mar 2025
Published online: 14 May 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article