Analysis of PG Admission in Arts and Science College using Data Mining Tools

Abstract

Information is the backbone of any Organization. To succeed in this situation, one must manage the information in the right amount and at right time. Data Mining is used to mine the new pattern or trends or rules from the unknown/ large amount of data / unpredictable data sets. Data Mining is used at diverse fields like Educational, Agriculture, Medical, Police Department, Research side, Information Technology side and Image processing etc., This Paper analyzes the mindset of Final year UG student about PG admission. For that 18 attributes and one class label collected from the final year UG students then applied on the Data Mining Tools like Weka Tool and Orange Tool. The data sets passed on the classification algorithms like J48, Naive Bayes, RandomForest and REPT Tree in Weka and Classification Tree, CN2, Naive Bayes and kNN of Orange Data Mining Tool. The Confusion matrix, Training and simulated Errors and Testing and Validation Results are obtained and tabulated. The Weka and Orange data mining tools classifiers performance are represented in the graphical form and its decision tree. From the decision tree, hidden rules are extracted, from which possible to determine factors which affects the PG admission. From the Weka tool, obtain attributes interest, jobavailability, feestat, colinfra and scholarship can predict the PG admission. In orange data mining tool, attributes interest to study, jobavailability, feestatus, scholarship, gender, pregovexam and colinfra from the original data set can predict the PG admission