The research was conducted to reveal the effect of LQ45 stock on the accuracy of stock price index fluctuations using the C4.5 algorithm with Correlation-Based Feature Selection (CFS) and Information Gain (IG) techniques. This study used the superior C4.5 algorithm using a combination feature selection technique between Correlation-based Feature Selection (CFS) and Information Gain in the hope of getting accurate results. Analysis conducted on the LQ45 index through various stages that include data collection, manual pre-processing, validation methods, process features, decision tree model result, and classification accuracy performance. The result of test revealed that the implementation of the C4.5 algorithm using correlation-based feature selection (CFS) and information gain techniques can be applied well to LQ45 stocks. The accuracy generated from the original data (without the selection feature) was 77.857%, while the addition of features to the combination of Correlation-Based Feature Selection (CFS) and Information Gain had a large influence on the results of increasing data accuracy from the accuracy of the original data by 77.857% to 78.333%. Thus, the C4.5 calculation process with the Correlation-based Feature Selection (CFS) feature selection technique alone cannot improve the accuracy level, while when combined with the Information Gain technique, the accuracy processing results will be better (higher).