Job classification: the application of feature selection techniques on graduates’ data / Muhammad Haziq Haikal Hisham, Mohd Azri Abdul Aziz and Ahmad Asari Sulaiman

• Main Authors: Hisham, Muhammad Haziq Haikal, Abdul Aziz, Mohd Azri, Sulaiman, Ahmad Asari
• Format: Article
• Language: English
• Published: UiTM Press 2023
• Subjects: Probes (Electronic instruments)
• Online Access: https://ir.uitm.edu.my/id/eprint/76356/1/76356.pdf; https://ir.uitm.edu.my/id/eprint/76356/; https://jeesr.uitm.edu.my/

n this paper, job classification is viewed as a process to classify or to recommend jobs to the graduates according to the criteria set. The purpose of this study is to compare three feature selection techniques on the graduates’ data to determine the relevant features in the job classification process for graduates. The experiment included three different feature selection techniques which are Analysis of Variance (ANOVA), Chi-squared test, and Recursive Feature Elimination (RFE). The dataset used for the experiment covered 12 graduates’ feature that are needed to be tested to determine the impact of each graduates’ feature on the result. The final feature ranking was listed for each of the feature selection techniques used and two common features among the rank lists had been found out as important features that affect job classification among graduates.