Development of a Method for Evaluating Quality of Education in Secondary Schools Using ML Algorithms

Machine learning algorithms may have very wide area of applications. In this paper we used machine learning algorithms to establish a method for evaluating the quality of education in secondary schools, depending on their past experience. The tool developed can be used for performance comparison between different schools and future score prediction. We collected and compared the results of almost 650 students from various regions of Nigeria to establish a relationship between their academic performance in internal and external exams. Internal exams are those conducted by their respective schools while external exams are those held by independent bodies, like WAEC and JAMB. We conducted a regression test on UTME (JAMB) scores and classification test on WASSCE (WAEC) scores. With simple but effective algorithms, we managed to reduce the mean squared error by %75 for regression model, and improved the prediction accuracy in classification by %35. Model development was done by using Python libraries. With a developed model, we compared performances of the schools from different regions in Nigeria. Results show that findings are acceptable and applicable for further use.