Feature Selection In Support Vector Machine And Random Forest Algorithms For The Classification Of Recipients Of The Smart Indonesia Program

Education is very important for everyone in order to provide intelligence and build a good soul in themselves. Indonesia has a high dropout rate. This is because the education level of the Indonesian people is still in the low category. So the government runs a Smart Indonesia Program (PIP) to help improve the quality of education. As in SMAN 1 Tualang. This research classifies prospective PIP recipients using Feature Selection Chi Square and Information Gain with SVM and Random Forest Algorithms. There are 1112 students with 26 selection attributes. This study concluded that the Random Forest algorithm with Feature Selection is better than the SVM algorithm with the highest accuracy value of 99%, while the SVM algorithm has the highest accuracy value of 98%. Feature selection that has a big influence in determining accuracy is Information Gain.