Intelligent prediction model for learners outcome forecasting in e-learning

Abstract

In e-learning environment, users are very much interested in predicting the outcomes and monitoring the learning process to verify their prediction. Traditional machine learning techniques includes objective prediction (quantitative measure with an abundance of data) and subjective forecasting (qualitative measure with small data) methods. However, these techniques may not be consistent in various situations. In this research paper, we present an intelligent prediction model for learners outcome forecasting approach, which helps facilitators and users discover more interesting knowledge information and predict the learning outcomes. A high level machine learning technique identifies partial similarities between learners time series data and categorize the data group in to various group based on their similarity computation. A modern visualization of the data categorization process helps us to understand the similarity between the time series information. Statistical measures evaluate the effectiveness of the proposed approach of categorization and testing their significance. Evaluation results show that our technique leads to relatively high accuracy in learners outcome prediction.