Prediction of MUET result based on KNN algorithm / Siti Fatimah Azzahra Hamrizan

Malaysian University English Test (MUET) has been used to measure students' English proficiency level before entering bachelor programmes in Malaysian universities. UiTM offers English language proficiency courses (ELC) for semester 1 to semester 3 diploma students to prepare them for MUET. The exam results from the ELC can be the features to help the students to know if they can pass the MUET exam and the lecturers can figure out which language skills need to be improved to help preparing the students to sit for MUET. The study aims to explore the K-Nearest Neighbour (KNN) algorithm in solving the MUET result prediction problem, to develop a prototype of MUET result prediction based on the KNN algorithm and to evaluate the accuracy of the KNN algorithm in MUET result prediction. The machine learning technique used to develop the prediction prototype is the KNN algorithm. The results show that the highest average accuracy was at 65.294% and the percentage error was at 25.490%. Some additional future works should be applied to improve the algorithm performance are to add more data, use other related results as the features and compare the KNN algorithm to other algorithm.