Comparison and Prediction of Data Mining Models to Determine the Classification of Family Planning Program User Status

Abstract

<p class="JRPMAbstractBodyEnglish">Data mining is the process of analyzing a sample to determine the best performing algorithm. An easy way to extract information or insights from large amounts of data is by using the techniques involved in data mining. There are several methods of classification which can be used to determine the level of a certain acuity. In Indonesia, family planning program is the most common program in the government to control population growth. A decision tree, a logistic regression, a naive Bayes model, and a gradient boosted model are used in this study. To perform classification family planning program User Status in Mangunharjo sub-district, the variable used is the wife's age, age of the youngest child, stages of prosperous family, and number of children. The training data comparison testing is 70:30. This study was tested by using the AUC value and t-test. The best value for accuracy is the Decision Tree algorithm with a percentage of 94.2% and an AUC value of 0.939. From the results of this test, it can be concluded that for a comparison of all tests performed on the dataset, the Decision Tree algorithm model can be said to be better than the other three algorithm models.</p>