Perbandingan regresi logistik biner dengan jaringan syaraf tiruan (studi pada klasifikasi status akreditasi Sekolah Menengah Pertama Provinsi DKI Jakarta)

Abstract

Classification is one of statistical technique in classifying data compiled systematically. One popular example of classification in the field of education is the accreditation status of the school. Accreditation status of schools affected by the standard of Teachers and Education Personnel (TOD). There are many statistical methods that can be used to solve classification problems, including binary logistic regression and Artificial Neural Network (ANN). The result of the calculations in studies using binary logistic regression gives an Area Under Curve (AUC) of 0.852 for training data and 0.834 for testing data. Calculation with a back propagation ANN for training data and testing data give an AUC of 0.911 and 0.899. Thus, the back propagation ANN can predict better than the binary logistic regression. The variables that significantly affect the classification of the accreditation status of the school is the number of teachers (X1) and the qualifications of chief administrative personnel (X6).