DEVELOPMENT OF MACHINE LEARNING BASED SOFTWARE FOR PREDICTIONS AND CLASSIFICATION OF STUDENTS’ ACADEMIC PERFORMANCE IN FEDERAL UNIVERSITIES, SOUTH-SOUTH, NIGERIA

Abstract

The study developed a software driven by machine learning algorithm for the Predictions and Classification of Students’ Academic Performance in Federal Universities, South-South Nigeria. The purpose of this study was to development of Machine Learning based software for Predictions and Classification of Students’ Academic Performance in Federal Universities, South-South Nigeria. The study also investigated the perceive usefulness, ease of use and social influence of the Machine Learning based software. The study was guided by five specific purposes of the study. Three research questions were raised. This study was carried out in Federal Universities, South-South Nigeria. Two designs were adopted in the study, Rapid Application Development model design and Descriptive Survey research design. The population of the study consisted of 9934 lecturers in Federal Universities, South-South Nigeria, during the 2023/2024 academic session. A sample size of 397 lecturers which consisted of 198 Examination Officers and 199 Staff Advisers were selected for the study. The sample of the study was drawn using simple random sampling technique. A researcher made instrument titled “Usability of Machine Learning based Software Questionnaire” (UMLBSQ) was used in collecting data. The instrument was face validated by three experts. The reliability coefficient of 0.82 was obtained using Cronbach alpha Statistics. Mean and standard deviation were used to answer research questions while independent t-test was used to test the hypotheses at 0.05 level of significance. The result revealed that the respondents perceived the system as useful and easy to use. It was therefore recommended among others that Students should be allowed to engage with the predictions and classifications to set achievable goals and improve study strategies for better academic outcomes.