Extreme Bound Analysis Based on Correlation Coefficient for Optimal Regression Model

Regression analysis is an important tool in statistical analysis, which there a demand of discovering essential independent variables among many other ones, especially case that huge number random variables. Extreme bound powerful approach to extract such called robust regressors. In this research, I propose so-called Regressive Expectation Maximization with RObust regressors (REMRO) algorithm as alternative method beside probabilistic methods for analyzing By the different ideology from methods, REMRO searches forming optimal regression model and sorts them according descending ordering given their fitness values determined by two proposed concepts local correlation global correlation. Local represents sufficient explanatories possible regressive models reflects independence level stand-alone capacity Moreover, can resist incomplete data because it applies (REM) into filling missing estimated based on expectation maximization (EM) algorithm. From experimental results, more accurate modeling numeric than traditional like Sala-I-Martin but cannot be applied nonnumeric yet research.