Development of Machine Learning Models to Predict Student Performance in Computer Literacy Courses
Year: 2018
Type: article
Abstract: This paper reports on a study carried out to build machine learning models for the purpose of predicting student performance in a second course of a two-course sequence, based on performance in various aspects of the first course’s exam. Detailed data is collected from the first course’s exam, which is a multiple choice exam, using an Optical Mark Recognition scanner. This data is then pre-processed and used to build machine learning models. Several machine learning models are experimented with, yielding excellent results for predicting Pass or Fail (greater than 93% accuracy), validating our hypothesis that our approach can be used to help students with preventative measures in order to increase pass rates in the courses concerned.
Authors: George Anderson, Oduronke T. Eyitayo
Institution: University of Botswana
Language: English
Subfield: Computer Science Applications
Field: Computer Science
Domain: Physical Sciences
SDG: Quality education
Open Access status: green
OpenAlex ID: W2932182669
