Student Performance Analysis using Machine Learning

Abstract

Predicting student data to improve instructor and learner more efficiently in teaching and learning. It also strengthens contact between administrators, teachers and learners and helps monitor the behavior of students at multiple levels, such as class assignments, workshops, internal assessments and final exams. This program was built across three fields. We are Learning, Psychology and Computer Science. Educational institutes are increasingly using educational systems in recent years to assess their performance in order to construct plans for further growth and future actions. Such activities concentrate on discovering and improving approaches that can improve student academic performance, indirectly helping institutes attract more new students and maintain older students, the algorithms used in these systems are known as Educational Data Mining or EDMs. The prediction of student performance is an important aspect of EDM, which is the main area of this research work. Predicting student performance is a process that focuses on inferring information to learn from the student performance data.It can provide accurate collection of data on learning activities, such as time-on-task and evaluation scores, allowing for useful progress estimates for both the student and the trainer. In order to improve their performance, early warning of "at risk students" can be obtained which can help trainers to increase their focus on them. This provides a better way of predicting student performance to improve their skills at an earlier stage. Thus, student performance prediction helps to easily adapt, personalize and interfere.