In the analyzed students’ educational data several parameters such as True Postive Rate, False Positive Rate and Classification Error were used as a yard stick in measuring the performance of both Kstar and BayeNet algorithms in mining the educational data. The performance investigation of the applied classifiers revealed hidden knowledge in the data set which was helpful in the re-calibration of the model to yield a higher precision of each of the classifier with minimal classification error.
展开▼
