Application of Decision Tree for Predictive Analysis of Student’s Self Satisfaction with Multivariate Parameters

Abstract

Decision trees are better known for automatic feature selection in predictive analysis. There is no need of variable transformation while processing data in decision tree as such trees work on selection of important variables as parent nodes and they act as split node. An attempt is made herein to enumerate decision tree over customized educational dataset. The work is based on the hypothesis that distinctiveness, routine and sensitivity related variables of students have close association together with self satisfaction and motivation. This paper demonstrates better feature selection using tree analytics. Experimental analytics were carried out with R software package.