FEATURE SELECTION BASED ON CHI SQUARE IN ARTIFICIAL NEURAL NETWORK TO PREDICT THE ACCURACY OF STUDENT STUDY PERIOD

A graduation prediction of student can be regarded as a nonlinear classification problem involving some academic parameters that has a goal to predict whether a student can graduate or not. One of the best classification methods which can be used to solve the problem is an Artificial Neural Network (ANN). Certainly, ANN needs a sufficient training data to obtain knowledge in order to classify the data since the inadequate training data can decrease the accuracy of the ANN. The accuracy can be increased by conducting a feature selection which becomes an input for ANN and eliminating the unimportant features which do not have correlation with the output of the ANN. In this research, the statistic approach using the Chi Square method was proposed to select the feature in student academic record data to be the input of ANN. The use of the Chi square succeeds to show which features having a significant influence towards the output of the ANN.