Predicting Students’ Inclination to TVET Enrolment Using Various Classifiers

Abstract

Technical and Vocational Education and Training (TVET) is an education system that delivers necessary information, skills, and attitudes related to work or self-employment. However, the TVET program is not preferred by most Malaysian students due to several factors such as students’ interest, parental influence, employers’ negative impression, facility in vocational institutions, inexperienced TVET instructors, and society’s negative perception. Consequently, it raises the issue of skilled workers shortage. The gravest threat will be far-reaching, pushing our economy into depreciation. Therefore, it is important to identify the students’ traits and interests before conducting further investigation to turn and thrive in this phenomenon. This study aims to utilise several classifiers (Decision Tree, Neural Network, Logistic Regression and Naïve Bayes) to predict students’ inclination to join TVET programmes. A total of 428 secondary school students from Kedah, Malaysia, are chosen as our survey respondents. The best classifier is determined according to the lowest misclassification rate. The findings revealed that the Decision Tree-based Gini Index with three branches prevail against other classifiers with a misclassification rate of 0.1938. Therefore, the classifier could act as a steer for the Kedah Department of Education (DOE), related parties, and the TVET agency in implementing effective strategies to enliven and inspire students to join TVET programs.

Hybrid Machine Learning Approach for predicting E-wallet Adoption among Higher Education Students in Malaysia

“…A decision tree is a tree-structured classifier where core nodes represent the characteristics of a dataset, branches represent the decision rules, and leaf nodes represent the outcomes (Ch'ng & Mahat, 2014;Hong et al, 2023). Classification and regression tree algorithm (CART) is a type of classification algorithm that uses Gini's impurity index to build a decision tree model (Breiman et al, 1984).…”