Regresi Logistik Biner dalam Penentuan Ketepatan Klasifikasi Tingkat Kedalaman Kemiskinan Provinsi-Provinsi di Indonesia

Abstract

This study aims to find out the variables that affect the poverty depth index of provinces in Indonesia as well as the level of classification produced by logistic regression analysis method. This analysis was chosen because the variables tied to this study had values of 0 and 1 (categorically). The variable tied to this research is the level of poverty depth of provinces in Indonesia in 2019. While the free variables are PPP (adjusted per capita expenditure), MYS (Mean Years School) and EYS (Expected Years of Schooling) of provinces in Indonesia in 2019.  The population used in this study is all provinces in Indonesia with sampling techniques is census because the population is less than 100. The samples in this study are the same as the population of all provinces in Indonesia. The results showed that MYS and EYS influenced the level of poverty depth with the accuracy of the overall prediction is 85.3% which indicates that there are still differences in classification results obtained from the original data with the logit regression model obtained.