Optimizing adaptive neuro-fuzzy inference system model based Chaotic Harris Hawks algorithm for stock prediction

Artificial intelligence (AI) has become increasingly prevalent in the financial sector, particularly for predicting stock market movements. Due to the efficient market hypothesis, forecasting stock prices remains a challenging task, as prices are primarily influenced by fluctuating supply and demand. This study introduces an optimized Adaptive Neuro-Fuzzy Inference System (ANFIS) integrated with a Chaotic Harris Hawks Optimization (ChHHO) algorithm to enhance the accuracy of stock price predictions. The ChHHO method improves search space exploration and mitigates the risk of convergence to local optima, thereby enhancing predictive performance. This study is motivated by the need to improve the precision of stock price predictions, which are often hindered by the non-linear and chaotic nature of financial data. By combining ANFIS with the chaotic Harris Hawks Optimization, we aim to develop a model that addresses these challenges effectively. The proposed ANFIS-ChHHO model is evaluated on EGX index stock data using metrics such as Root Mean Square Error (RMSE), Mean Absolute Error (MAE), Standard Deviation (SD), and Theil's U. Results demonstrate that the ANFIS-ChHHO model outperforms traditional methods in prediction accuracy.