DISCOVER THE ATTRIBUTES AFFECTING GRADE OF UNIVERSITY STUDENT THROUGH FEATURE SELECTION METHODS

Abstract

There are many techniques to choose subset features. This stage is the preprocessing of the data mining task. The process of choosing a subset of features from the original collection of attributes is known as feature selection. Obtaining relevant information to predict characteristics affecting the estimation of the CGPA of graduates of private universities is a hard task. The experiments were conducted on four datasets. Two methods are used in order to choose the features. They are CfsSubsetEval and Correlation Ranking. Correlation Ranking uses 0.1 as the threshold for feature selection. The results indicate that there is a relationship between a student's grade in the previous qualification and his grade when he graduates from the university.