Early Forecasting of Stock Marketing Price Using SVM Classifier by Comparing with Decision Tree for Improved Accuracy

The goal of this study was to evaluate how accurately and reliably Support Vector Machine (SVM) and Decision Tree (DT) classification algorithms can predict stock market trends. For this analysis, we used a dataset of 570 stock market records. The study aimed to compare these two algorithms to determine which one performs better in predicting stock market movements. To conduct this evaluation, we set up a framework that involved using both the Support Vector Machine and Decision Tree classifiers on the stock market dataset. We calculated the sample size needed for this analysis as 55 records per group, using G power analysis. The parameters for this calculation included an alpha value of 0.05, a beta value of 0.5, a 95% confidence level, and an 80% pre-test power, with an enrollment ratio of 1. The results showed that the SVM classifier achieved an accuracy of 59.76% when applied to the dataset, whereas the Decision Tree classifier achieved an accuracy of 51.85%. The statistical analysis revealed a significant difference between the two algorithms, with a p-value of 0.0184, which is less than the threshold of 0.05.