PERBANDINGAN METODE REGRESI ROBUST ESTIMASI M DENGAN PEMBOBOT TUKEY BISQUARE DAN ESTIMASI LTS DALAM MENGATASI OUTLIER (STUDI KASUS: FAKTOR YANG MEMPENGARUHI INDEKS PEMBANGUNAN MANUSIA DI JAWA TENGAH PADA TAHUN 2024)

Edia Hestiningtias, NIM.: 21106010068 (2025) PERBANDINGAN METODE REGRESI ROBUST ESTIMASI M DENGAN PEMBOBOT TUKEY BISQUARE DAN ESTIMASI LTS DALAM MENGATASI OUTLIER (STUDI KASUS: FAKTOR YANG MEMPENGARUHI INDEKS PEMBANGUNAN MANUSIA DI JAWA TENGAH PADA TAHUN 2024). Skripsi thesis, UIN SUNAN KALIJAGA YOGYAKARTA.

Preview	Text (PERBANDINGAN METODE REGRESI ROBUST ESTIMASI M DENGAN PEMBOBOT TUKEY BISQUARE DAN ESTIMASI LTS DALAM MENGATASI OUTLIER (STUDI KASUS: FAKTOR YANG MEMPENGARUHI INDEKS PEMBANGUNAN MANUSIA DI JAWA TENGAH PADA TAHUN 2024)) 21106010068_BAB-I_IV-atau-V_DAFTAR-PUSTAKA.pdf - Published Version Download (7MB) | Preview
	Text (PERBANDINGAN METODE REGRESI ROBUST ESTIMASI M DENGAN PEMBOBOT TUKEY BISQUARE DAN ESTIMASI LTS DALAM MENGATASI OUTLIER (STUDI KASUS: FAKTOR YANG MEMPENGARUHI INDEKS PEMBANGUNAN MANUSIA DI JAWA TENGAH PADA TAHUN 2024)) 21106010068_BAB-II_sampai_SEBELUM-BAB-TERAKHIR.pdf - Published Version Restricted to Registered users only Download (9MB) | Request a copy

Abstract

This study aims to analyze the effectiveness of robust regression methods in handling outliers in modeling the Human Development Index (HDI) in Central Java Province in 2024. The dependent variable used is the HDI (Y ), while the independent variables include Expected Years of Schooling (X1), Mean Years of Schooling (X2), Life Expectancy at Birth (X3), and Real Per Capita Expenditure (X4), based on regency and city data in Central Java Province in 2024. The analysis was conducted using the Ordinary Least Squares (OLS) method, M-estimation with Tukey Bisquare weighting, and the Least Trimmed Squares (LTS) estimation, following classical assumption testing and outlier detection using leverage, DFFITS, Cook’s Distance, and R-Student. The results indicate that the OLS method produces the largest Residual Standard Error (RSE) of 0,1443, while M-estimation with Tukey Bisquare weighting yields a smaller RSE of 0,1171. The LTS method demonstrates the best performance, with the smallest RSE of 0,05997 and the highest coefficient of determination of 0,9996. Therefore, the LTS method is the most optimal approach for addressing outliers in HDI data and produces a more stable and accurate regression model.

Share this knowledge with your friends :

Actions (login required)

View Item