An efficiency comparison of data classification with data mining techniques in study plan prediction for upper secondary admission in Suranari Wittaya School
Abstract:
In this research, the results of comparing the performance of data classification techniques using 3 data mining techniques, namely Decision Tree, Random Forest, and k-Nearest Neighbor (k-NN), were presented. The algorithm with the highest performance was selected to be used as a study plan prediction system for upper secondary admission at Suranari Wittaya School. The study utilized a selection study plan result dataset consisting of 1,772 records from former 9th-grade students during the academic years 2020-2022. The research findings demonstrate that the Random Forest algorithm exhibited superior efficiency, achieving an accuracy rate of 80.81%, precision rate of 91.09%, and recall rate of 69.88%. The Decision Tree algorithm and the k-Nearest Neighbor algorithm followed, achieving accuracy rates of 73.31% and 70.20%, precision rates of 71.21% and 66.98%, and recall rates of 59.14% and 52.44%, respectively.