Student Performance Classification using Adap...

Student Performance Classification using Adaptive DNN with SVM Approach

Author Name : Soni Darshan Kumar Babulal, Prof. Dr. Binod Agrawal

ABSTRACT:  Predicting student performance based on their academic score is critical for any academic organization; a number of strategies can be used to improve or maintain student performance during their studies. The academic performance in this research is measured based on marks obtained by students of different colleges. This research aims to examine the accuracy rate of the student performance prediction system, which is designed using the concept of data mining with a machine learning approach. The foremost step of this research is to collect data from the database and then cleaned unstructured data into stricture form by using the concept of Cosine similarity with the K-mean clustering approach. After that, the clustered data is used to train the Deep Learning Neural network (DNN), and also, the cross-validation of the result has been performed using a Support Vector Machine (SVM) approach. At last, the performance parameters are measured. The results indicate that the SVM with the DNN approach performs well and provides better prediction accuracy to analyze the performance of students.