Placement chance prediction: Clustering and classification approach

Abstract

Educational data mining is an area wherein a combination of techniques, such as data mining, machine learning and statistics, is applied on educational data to get valuable information. The purpose of this paper is to help prospective FAD students in selecting or choosing the right undergraduate course, viz., accessory design, fashion design, textile, fashion communication, etc., based on the entrance exam ranking for admission to the UG course. A clustering and classification approach is applied to solve the placement chance prediction problem. Two classification algorithms, viz., decision tree and Naïve Bayes and a clustering algorithm K-means are applied on the same data set. Algorithms applied are compared and it was found that clustering algorithm K-means predicts better in terms of precision, accuracy, and true positive rate. This work will help students in selecting the best course suitable for them that ensures best placement chance.