%0 Thesis
%9 Skripsi
%A Laili Mahfuzhotul Hasanah, NIM.: 21106010052
%B FAKULTAS SAINS DAN TEKNOLOGI
%D 2025
%F digilib:71773
%I UIN SUNAN KALIJAGA YOGYAKARTA
%K Robust, Least Trimmed Squares, M, Andrew, RSE
%P 142
%T PERBANDINGAN METODE LEAST TRIMMED SQUARE (LTS) DAN ESTIMASI M PEMBOBOT ANDREW DALAM REGRESI ROBUST (STUDI KASUS: FAKTOR KESEJAHTERAAN MASYARAKAT PADA INDEKS PEMBANGUNAN MANUSIA 2023)
%U https://digilib.uin-suka.ac.id/id/eprint/71773/
%X Multiple regression analysis aims to analyze the relationship between  two or more variables. One of them is by using the Ordinary  Least Squares (OLS). OLS aims to minimize errors that are not resistant  to outliers and cause bias and inaccuracy. Robust regression is one  alternative method to overcome this. Robust regression compares two  estimates: M with Andrew weights and Least Trimmed Squares (LTS).  The M estimate is designed to minimize the influence of data with large  residuals by replacing the error square function with a loss function that  is more resistant to outliers. Meanwhile, the LTS method uses the principle  of robust regression fitting by minimizing the smallest residual  squares, i.e., by excluding the largest residuals considered outliers. The  Human Development Index is one of the important indicators of a country’s  progress, which is influenced by factors such as health, education,  and the economy. This study aims to analyze the influence of factors related  to community welfare, including expected years of schooling, average  years of schooling, labor force, labor force participation rate, and  the number of public hospitals on the Human Development Index. The  results of the study show that OLS, M estimation, and LTS have RSE  values of 1.173, 0.9701, and 1.00035, respectively. M estimation has  the lowest RSE value compared to the other two estimations. This indicates  that M estimation is the most effective in producing a model with  better predictions in overcoming outlier and multicollinearity problems
%Z Sri Utami Zuliana, S.Si., M.Sc., Ph.D.