OPTIMISASI ANALISIS REGRESI LOGISTIK LASSO MENGGUNAKAN BAYESIAN INFORMATION CRITERION (BIC) (STUDI KASUS : FAKTOR PENYEBAB PERCERAIAN DI INDONESIA TAHUN 2024)

Mutiara Nur Amalina, NIM.: 22106010089 (2026) OPTIMISASI ANALISIS REGRESI LOGISTIK LASSO MENGGUNAKAN BAYESIAN INFORMATION CRITERION (BIC) (STUDI KASUS : FAKTOR PENYEBAB PERCERAIAN DI INDONESIA TAHUN 2024). Skripsi thesis, UIN SUNAN KALIJAGA YOGYAKARTA.

Preview	Text (OPTIMISASI ANALISIS REGRESI LOGISTIK LASSO MENGGUNAKAN BAYESIAN INFORMATION CRITERION (BIC) (STUDI KASUS : FAKTOR PENYEBAB PERCERAIAN DI INDONESIA TAHUN 2024)) 22106010089_BAB-I_IV-atau-V_DAFTAR-PUSTAKA.pdf - Published Version Download (6MB) | Preview
	Text (OPTIMISASI ANALISIS REGRESI LOGISTIK LASSO MENGGUNAKAN BAYESIAN INFORMATION CRITERION (BIC) (STUDI KASUS : FAKTOR PENYEBAB PERCERAIAN DI INDONESIA TAHUN 2024)) 22106010089_BAB-II_sampai_SEBELUM-BAB-TERAKHIR.pdf - Published Version Restricted to Registered users only Download (11MB) | Request a copy

Abstract

Divorce is a social issue influenced by various contributing factors and shows different patterns across provinces. This study aims to classify divorce levels in Indonesia in 2024 using LASSO logistic regression with optimal parameter selection based on the Bayesian Information Criterion (BIC). The data used in this study are secondary data from the Statistics Indonesia (BPS), covering 34 provinces in Indonesia. The response variable was categorized into two groups, namely low and high divorce levels, based on the divorce rate per 1,000 population. Meanwhile, the predictor variables consisted of 13 factors causing divorce. The analysis began with the development of an initial logistic regression model, followed by multicollinearity testing, variable standardization, the application of the LASSO method, and the selection of the optimal λ value using BIC. The results showed that the initial logistic regression model experienced multicollinearity, as all variables had VIF values greater than 10. Through the LASSO-BIC method, the optimal value obtained was λBIC = 0.003914, with a minimum BIC value of 26.77. The final model retained six variables, namely gambling, abandonment by one spouse, imprisonment of one spouse, polygamy, physical disability/illness, and continuous disputes or conflicts. Compared with the initial model, LASSO-BIC produced a simpler model, reducing the number of parameters from 14 to 7 and decreasing the BIC value from 66.055 to 26.770.

Share this knowledge with your friends :

Actions (login required)

View Item