![]() | Only 14 pages are availabe for public view |
Abstract This thesis aims to discover chances of a superiority of the fourth year students, English section, from 2009 to 2013. The technique of classification is applied, based on historical data reserved in database and records of student affairs. This study helps the faculty administration to use modern techniques to achieve a quick and easy patterns for evaluating and understanding the behavior of a student in the light of using the outcome of mining tasks in practical life. The study included (261) instances and (16) variables, in addition to four classification techniques: Naïve Bayes, K-Nearest Neighbor, MLP Artificial neural network, and the j48 decision tree. It is carried out through WEKA program. As a result, the Naïve Bayes classifier is the most accurate algorithm which gives the lowest error rate in addition to the highest accuracy from the ROC curve. |