A Comparative Study of Linear Regression and Regression Tree

Abstract

In data mining, prediction modeling is a technique used for finding a mathematical correlation between a dependent variable and various independent variables or predictor variables. Generally, the linear regression describes the correlation between multiple independent or predictor variables and one response or dependent variable. In machine Learning, the CART model builds for classification and regression purpose when the decision tree is applied for classification purpose then it refers to classification tree and if it is applied for regression then it may refer to regression tree. A decision tree is also known as the Regression tree. In this paper, we are concentrated on the concepts of Linear regression and Regression tree. A dataset is incorporated from UCI (Machine learning Repository) for this research work. The objective of the study is to distinguish the results obtained from Linear regression and regression tree. In the end, the decision tree gives better results than linear regression and the final predictive model is selected on a minimum mean sum of the square.