%0 Thesis
%9 Skripsi
%A Zainab Yumna Shalihah, NIM.: 20106010030
%B FAKULTAS SAINS DAN TEKNOLOGI
%D 2026
%F digilib:75390
%I UIN SUNAN KALIJAGA YOGYAKARTA
%K Multikolinearitas, Regresi LASSO, Algoritma LARS, Akaike  Information Criterion, Indeks Pembangunan Manusia
%P 86
%T OPTIMISASI ANALISIS REGRESI LASSO DENGAN ALGORITMA LARS MENGGUNAKAN AKAIKE INFORMATION CRITERION (AIC) (STUDI KASUS : INDEKS PEMBANGUNAN MANUSIA DI PROVINSI JAWA TENGAH TAHUN 2023)
%U https://digilib.uin-suka.ac.id/id/eprint/75390/
%X Multicollinearity is a common problem in multiple linear regression due to the high correlation between independent variables, which can cause instability in parameter estimates, increase coefficient variance, and decrease the statistical significance of the model. This study aims to address the problem of multicollinearity by applying the Least Absolute Shrinkage and Selection Operator (LASSO) method with the Least Angle Regression (LARS) algorithm and selecting the optimal  penalty parameter using the Akaike Information Criterion (AIC). The analysis process begins with data standardization and multicollinearity identification using the Variance Inflation Factor (VIF), followed by LARS-based LASSO regression estimation. The penalty parameter (λ) value is determined based on the minimum AIC value to achieve an optimal balance between model accuracy and parameter complexity. This method is applied to the 2023 Human Development Index data of  Central Java Province by district/city. The results of the study show that LASSO regression is able to reduce the impact of multicollinearity through effective variable selection, resulting in a simpler, more stable, and easier to interpret model than conventional linear regression.
%Z Sri Utami Zuliana, S.Si., M.Sc., Ph.D.