Citation

Amir Hamzah, Nur Asyiqin and Ab Ghani, Hadhrami and Malik, Nurmasliza and Thman, Muhammad Hanafi and Ibrahim, Mohd Hakimi Aiman (2025) Predicting Career-Driven Academic Pathways Using Hybridized Machine Learning Strategies. In: 2025 IEEE International Conference on Communication, Networks and Satellite (COMNETSAT), 11-13 December 2025, Padang, Indonesia.

Text 9.pdf - Published Version Restricted to Repository staff only Download (1MB)

Abstract

Selecting a suitable study route during secondary education represents a pivotal milestone, as it directly influences both academic performance and future employability. Yet, many learners make these decisions based on limited information or generic advice, often leading to suboptimal choices. This work proposes a data-oriented approach that employs hybrid machine learning methods to provide personalized pathway recommendations. Student records comprising grades, personal attributes, and career aspirations were collected, cleaned, and analyzed. Multiple hybrid configurations were proposed and tested, including a stacking ensembles, a deep and wide network and a Tabnet model. Results highlight that the stacking ensemble achieved the highest accuracy among all, affirming the robustness of combining diverse base learners. The wide and deep network achieve quite a satisfactory accuracy, highlighting its capability to effectively handle both linear and non-linear feature relationships.

Item Type:	Conference or Workshop Item (Paper)
Uncontrolled Keywords:	Educational recommender systems, hybrid ma chine learning, predictive analytics, student guidance, study pathway decision-making
Subjects:	Q Science > QA Mathematics > QA71-90 Instruments and machines
Divisions:	Faculty of Engineering and Technology (FET)
Depositing User:	Ms Suzilawati Abu Samah
Date Deposited:	20 Apr 2026 02:08
Last Modified:	20 Apr 2026 02:08
URII:	http://shdl.mmu.edu.my/id/eprint/15750

Downloads

Edit (login required)