Handling Outlier In The Two Ways Table By Using Robust Ammi And Robust Factor

Abstract

Robust Regression is a regression methods were used to analyze the data that contains some outliers. This regression is a statistical model were easy to influenced to small changes in the data. This method is often used for data analysis on additive model. In a modeling of two ways table, it has been known AMMI (Additive Main and Multiplicative Interaction) which can be used to analyze the stability of genotypes at several different environments by combining the additive model of main effect and multiplicative model of interaction. AMMI models used for data with normally distribution. AMMI models will against the same problem if there are outliers in two ways table. Because of that problem, it is necessary a robust method in decompotition of interaction matrix including robust SVD and Robust PCA. This study analyzed data on two-way tables that contain contain outliers by using approach of robust SVD and approach of robust PCA. The results of this study on both methods will be compared the goodness of model through the comparison of biplot of each model.