Klasifikasi Penerima Program Indonesia Pintar Menggunakan Algortima Naïve Bayes Dan Random Forest

Abstract

Abstract Getting an education is certainly inseparable from the educational problems that are often faced by someone, namely, the cost of education. The Smart Indonesia Program (PIP) is part of President Joko Widodo's policy for poor and vulnerable families to get a good education for their children at no cost. The age of children covered by education costs is from the age of 6 to 18 years. With the Smart Indonesia Program, it is hoped that the dropout rate can drop drastically. The selection of PIP scholarships is carried out by the school. The decision-making process still uses data input carried out by school operators through the Dapodik application, so that many PIP recipients' decision-making is not on target. Subjectivity can occur in decision making as a result of inaccurate data. This study aims to classify recipients of the Smart Indonesia Program (PIP) using machine learning techniques with nave Bayes and random forest methods.This Smart Indonesia Classification Program uses machine learning techniques with the Nae Bayes algorithm and Random Forest. Sample data or secondary data comes from SMK Negeri 1 Cirebon which is used to predict the beneficiaries of the Smart Indonesia Program to facilitate decision making. The dataset includes the attributes of parents' occupations, total parental income, number of dependents, father's income, mother's income, KIP recipients, and families receiving other social assistance such as the Family Hope Program (PKH), Prosperous Family Card (KKS) and Social Protection Card ( PPP). It is expected that the accuracy results from the classification of the recipients of the Smart Indonesia Program (PIP) are 99.96% using the nave Bayes algorithm, while the accuracy results using random forest are 78.42%. against this dataset, it turns out that the nave Bayes algorithm is 21.54% better in accuracy than the random forest algorithm.