Educational Data Analysis and Classification using Deep Learning Techniques

In today's cutthroat society, it is essential for an institution to predict student performance, classify people according to their talents, and work to improve their performance in upcoming exams. To increase academic achievement, students should be instructed to focus their efforts in a particular area well in advance.This research propose novel method in educational data analysis based on academic student data performance using deep learning techniques by feature extraction as well as classification. Input has been collected as academic student performance data and processed for noise removal, normalization and dimensionality reduction. The processed data features has been extracted utilizing kernel quantum based reward Q-neural network and classified using ensemble VGG-19 with encoder_ convolutional architecture. Various student academic performance data are subjected to experimental analysis in terms of accuracy, precision, recall, F-1 score, RMSE, and MAP. Proposed method achieved 99% accuracy, 98% precision, 99% recall, a 98% F-1 score, 0.002% RMSE, and 99% MAP.