<![CDATA[The Jakarta Composite Index (IHSG) is a key indicator that reflects the performance of the stock market in Indonesia. It is often used by investors, analysts, and decision-makers to assess economic conditions and make investment decisions. However, the fluctuating and dynamic nature of the stock market makes predicting the IHSG a significant challenge. This study compares the effectiveness of Neural Network (NN) and Support Vector Machine (SVM) with optimization methods such as Particle Swarm Optimization (PSO) and Evolutionary Algorithm (EVO) in predicting stock prices. The results show that the combination of SVM with EVO provides the best prediction accuracy with the lowest error values (RMSE: 0.07, MAE: 0.09, MSE: 0.004). In contrast, NN with PSO and EVO showed higher prediction errors, indicating lower accuracy levels. These findings highlight the potential of optimization methods in enhancing the performance of stock prediction models, with SVM+EVO being the most effective combination.]]>
Copyrights © 2024
