Machine Learning Based Performance Evaluation System Based On Multi-Categorial Factors

Educational data mining is a focus of research reformed every growing year. The valuable data retrieved from various mining techniques are applied and implemented according to their own visions and objectives for developing specific technical systems to discover configurations and patterns from educational statistics. The performance evaluation of student model with advantage of behavioural (interactive) features is projected in this study paper and predicts student grade and level. The learner's interactions are directly connected with the category of feature. Few classification algorithms like Artificial Neural Network (ANN) and lib SVM are applied to analyze the impact of these categories on learners' academic performances. Results reflect the vigorous relationship between student's behaviour and his academic accomplishments. It raises the better results were observed with dissimilar classification methods involving interactive features while comparing classification accuracy with the identical data and eliminating such features.