Open AccessHoldout Validation for Comparison Classfication Naïve Bayes and KNN of Recipient Kartu Indonesia Pintar
Author(s) -
Firman Tempola,
R Rosihan,
Robiatul Adawiyah
Publication year - 2021
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/1125/1/012041
Subject(s) - naive bayes classifier , computer science , process (computing) , data mining , artificial intelligence , bayes' theorem , field (mathematics) , selection (genetic algorithm) , cross validation , feature selection , machine learning , mathematics , bayesian probability , support vector machine , pure mathematics , operating system
Kartu Indonesia Pintar (KIP) is one of the leading programs in the field of Education in the current government era. At present in the provision of smart Indonesian cards often occur not on target. The selection process for smart Indonesian cards is not transparent. For this reason, a transparent selection process is needed so that it does not cause jealousy among students. One way to make the selection process more objective is by applying existing classification methods to data mining. The study applies two methods to see the comparison, namely naïve bayes and k-nearest neighbor. In addition, the algorithm validation stage uses holdout validation. The purpose of algorithm validation is done so that each data has the same opportunity in the training and testing process, in this study also applied 150 dataset, the results showed that the system with no validation algorithm naïve Bayes accuracy is better that is an average accuracy of 85.66% and an average accuracy of k-nn 84.89%. However, if algorithm validation is applied, the k-nn accuracy is better at 88.7% compared to naïve bayes which is only 81.3%.