IAAPF: A Framework with Intelligent Methods and Feature Selection for Academic Achievement Prediction Using Machine Learning

Artificial Intelligence (AI) has become widely used emerging technology for solving real world problems. With learning based approach, machine learning (ML) models are capable of learning from data and predict future possibilities. This kind of application of ML in education domain has far reaching benefits. The problem identified in academic institutions is that there is lack of AI based solutions to predict students’ academic achievement. Traditional approaches are efficient but fail in achieving more comprehensive analysis of data due to the bulk of historical data going unprocessed. There is need for AI based solution to help mentors to predict student performance in advance and take necessary steps. However, it is a challenging problem to be addressed due to complexities involved in arriving at decision making. To solve this problem, in this paper, we proposed a framework known as Intelligent Academic Achievement Prediction Framework (IAAPF). This framework has underlying ML models and Long Short Term Memory (LSTM) based deep learning model for students’ academic performance prediction. The framework predicts students’ academic performance early in terms of “low” or “high” which leads to required intelligence to mentors. We proposed two algorithms to realize IAAPF.