Predictive Analysis for Student's Performance Evaluation

Abstract

These days, the instructional learning methods are not restricted to the traditional strategies, which used to be ten years prior. The 21st-century students need to be outfitted with information and abilities to establish fruitful and long-lasting learners. Innovation made it conceivable to attempt new learning strategies. So, with the development of technology and the necessary technology being more and more affordable and accessible to the general public, learning online is gaining popularity. Our exploration objective was to examine and investigate the Online Activity information to acquire significant knowledge of educators and their educating designs and eventually devise a prediction model to anticipate the outcome dependent on their action inside the learning time frame. Subsequently, Steps like data selection, data generation, data structure, feature engineering, feature selection were applied. Then, distinctive ML algorithms like CATBOOST, XGBOOST were applied for the prediction. This paper outlines the methods used to predict student's final results by various Machine Learning algorithms.