Perbandingan Metode Least Trimmed Squares dan Penduga-S dalam Mengatasi Data Pencilan dengan Simulasi Data

Abstract

Regression analysis was used to determine the relationship between variables. One of method parameter estimator in the regression model is ordinary least squares (OLS). In this study used four groups of data models with different outlier layout with five repetitions of each model. Then, this paper aims to compare the two methods, namely robust regression of least trimmed squares (LTS) and Sestimators. In the outliers are located amid robust regression line regression LTSestimators provides better results than the S-estimators, otherwise S-estimators is better at outliers are located end. The comparison criteria using the average squared residual.