Penerapan Model Regresi Logistik Biner pada Analisis Faktor yang Mempengaruhi Pengangguran Terbuka di Sumatera Utara

Abstract

Unemployment is a serious issue in Indonesia, leading to internal inequality and income distribution disparities. North Sumatra is among the ten provinces with the highest unemployment rates in Indonesia. The objective of this study is to develop a mathematical model based on factors influencing open unemployment in North Sumatra. The method applied is binary logistic regression using a categorical response variable with two categories: high and low. The maximum likelihood estimation method was used to estimate the parameters. Out of the 12 predictor variables studied, only 4 had a significant impact on the model, resulting in g(x)=−0,727−3,542x2(2) +2,166x4(2) + 3,135x5(2)−2,521x8(2). Based on the odds ratio calculation, the variable with a significant influence is the primary employment sector, with a value of 22.99. The impact of the response variable on the predictor variables in the model formed is 72.6%, with a classification accuracy of 86.1%.