Subitha Sivakumar ¹, Sivakumar Venkataraman ¹, Rajalakshmi Selvaraj ²

Affiliations
1 Faculty of Computing, Botho University, Botswana and Department of Information Systems, BIUST, Gaborone, BW
2 Department of Information Systems, BIUST, Gaborone, BW

Abstract

Background/Objectives: Educational Data mining is an emerging interdisciplinary research area that deals with the development of methods to explore data originating in an educational context. The objective of this work is to identify relevant attributes from socio-demographic, academic and institutional data from undergraduate students at the university located in India and develop an improved decision tree algorithm based on ID3 which can able to predict whether the students continue or drop their studies. Methods/Statistical Analysis: The traditional ID3 algorithm is improved by using Renyi entropy, Information gain and Association Function and the model generated by improved decision tree algorithm may be beneficial for university administrators to create guidelines and policies related to raise the enrollment rate in university and to take precautionary and advisory measures and thereby reduce student dropout. It can also used to find the reasons and relevant factors that affect the dropout students. Findings: Experimental results proved that improved decision tree algorithm provides better prediction accuracy in educational data than that of traditional classification algorithms in the literature. Improvements/Applications: Improved decision algorithm was proposed that enhances the ability to form decision trees and thereby to prove that the classification accuracy of improved decision algorithm on educational dataset is greater.

Keywords

Association Function, Decision Tree, Educational Data Mining, Information Gain, ID3, Renyi Entropy

Full Text

   

Sivakumar Venkataraman ¹, Subitha Sivakumar ¹, Rajalakshmi Selvaraj ²

Affiliations
1 Faculty of Computing, Botho University, Botswana and Department of Information Systems, BIUST, BW
2 Department of Information Systems, BIUST, BW

Abstract

Background/Objectives: A novel feature selection framework using minimum variance method is proposed. The purpose of the proposed method is to reduce the computational complexity, reduce the number of initial features and increase the classification accuracy of the selected feature subsets. Methods/Statistical Analysis: The clusters are formed using minimum variance method. The process must be repeated for different pairs of records and voting is done on the different sets of cluster pairs. The cluster pair which has the maximum number of votes is chosen and the highest priority member is chosen from each cluster using information gain and removing the remaining attributes, thus reducing the number of attributes. Findings: The proposed feature selector is evaluated by comparing it with existing feature selection algorithms over 9 datasets from UCI and WebKb Datasets. The proposed method shows better results in terms of number of selected features, classification accuracy, and running time than most existing algorithms. Improvements/Applications: A new feature selector using minimum variance method is implemented and found that it performs better than the popular and computationally expensive traditional algorithms.

Keywords

Classification, Data Mining, Dimensionality Reduction, Feature Selection, Information Gain, Minimum Variance Method

Full Text