Article Per Year (5 Year)
p-Index From 2020 - 2025
0.444
P-Index
This Author published in this journals
All Journal Jurnal Teknologi Informasi dan Ilmu Komputer International Journal of Electrical, Computer, and Biomedical Engineering (IJECBE)
Faza, Muhammad Naufal
Unknown Affiliation
Computer Science & IT Electrical & Electronics Engineering Engineering Materials Science & Nanotechnology Medicine & Pharmacology

Published : 2 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 2 Documents
Search

 1 
Defying Data Scarcity: High-Performance Indonesian Short Answer Grading via Reasoning-Guided Language Model Fine-Tuning Faza, Muhammad Naufal; Purnamasari, Prima Dewi; Ratna, Anak Agung Putri
International Journal of Electrical, Computer, and Biomedical Engineering Vol. 3 No. 3 (2025)
Publisher : Universitas Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.62146/ijecbe.v3i3.148

Abstract

Automated Short Answer Grading (ASAG) is crucial for scalable feedback, but applying it to low-resource languages like Indonesian is challenging. Modern Large Language Models (LLMs) severely overfit small, specialized educational datasets, limiting utility. This study compares nine traditional machine learning models against two fine-tuning strategies for Gemma-3-1b-it on an expanded Indonesian ASAG dataset (n=220): (a) standard fine-tuning predicting only scores, and (b) a proposed reasoning-guided approach where the model first generates a score rationale using knowledge distillation before predicting the score. The reasoning-guided model (Gemma-3-1b-ASAG-ID-Reasoning) achieved state-of-the-art performance (QWK 0.7791; Spearman’s 0.8276), significantly surpassing the best traditional model in this study (SVR, QWK 0.6952). This work advances foundational LSA-based approaches for this task by introducing a more robust methodology and evaluation framework. Crucially, standard fine-tuning (Gemma-3-1b-ASAG-ID) suffered catastrophic overfitting (QWK 0.7279), indicated by near-perfect training but poor test scores. While the reasoning-guided LLM showed superior accuracy, it required over 35 times more inference time. Results demonstrate that distilled reasoning acts as a powerful regularizer, compelling the LLM to learn underlying grading logic rather than memorizing pairs, establishing a viable method for high-performance ASAG in data-scarce environments despite computational trade-offs.
Implementasi Algoritma Genetika dalam optimasi Performa Truk Sampah Menggunakan Aplikasi Trash Queen Nugroho, Arief Kelik; Faza, Muhammad Naufal
Jurnal Teknologi Informasi dan Ilmu Komputer Vol 10 No 1: Februari 2023
Publisher : Fakultas Ilmu Komputer, Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.25126/jtiik.2023105193

Abstract

Kebutuhan akan kendaraan dapat dibilang sebagai salah satu kunci dari berjalannya ekonomi dunia. Akan tetapi, efisiensi desain spesifikasi kendaraan masih menjadi topik hangat di kalangan desainer otomotif. Hal ini karena meskipun ada jutaan kendaraan di seluruh dunia, tidak semua desain kendaraan dapat menunjukkan performa yang optimal di segala keadaan. Aspek spesifikasi kendaraan yang memiliki pengaruh paling besar antara lain daya mesin, ukuran ban, dan berat kendaraan. Oleh karena itu, dibutuhkan suatu optimasi desain spesifikasi kendaraan yang hanya membutuhkan faktor ukuran ban dan berat kendaraan, di mana berat kendaraan tersebut dapat direpresentasikan oleh dua faktor besar, yaitu kapasitas kargo dan kapasitas bahan bakar. Dalam penelitian ini ditelusuri kemungkinan optimasi spesifikasi truk sampah berdasarkan faktor-faktor tersebut dalam tujuannya mengumpulkan sampah dan kembali ke sentra pengumpulan sampah tanpa kehabisan bahan bakar menggunakan aplikasi Trash Queen. Trash Queen memanfaatkan algoritma genetika untuk menjalankan simulasi secara berulang-ulang hingga didapatkan solusi yang optimal. Truk sampah yang kehabisan bahan bakar tanpa mampu mengantarkan sampah akan dianggap gagal karena tidak mampu mencapai tujuannya. Pada penelitian ini, ditemukan bahwa dalam 30 generasi fitness terbaik tiap generasi telah naik sebanyak 68.4% dengan trend ukuran ban yang makin kecil, kapasitas bahan bakar yang makin kecil, dan kapasitas kargo yang makin besar. AbstractThe need for vehicles can be said to be the key in the continuity of the world’s economy. Even so, the efficiency of vehicle design specifications remain a hot topic among automotive designers. This is caused by the many millions of vehicles all across the globe, yet not all of them are able to perform to factory standards at optimal efficiency due to variations in different situations. The parts that have the most significant roles in a vehicle’s specification with respect to efficiency includes engine power, wheel size, and the weight of the vehicle itself. That is why a design optimization where it only accounts for the easier to find parts of a vehicle’s specification is needed, the wheel size and the weight of the vehcile which can be represented by two major factors: cargo capacity and fuel capacity. This research aims to explore the possibilities of optimizing a trash truck’s specifications in its conquest to collect trash and return to a trash collecting centre to deliver them without running out of fuel using the Trash Queen app based on those factors. Trash Queen utilizes genetic algorithm to run simulations repeatedly until an optimal solution is obtained. Trash trucks that run out of fuel before being able to deliver any trash will be considered as failed trucks due to being unable to accomplish their set goal. In this research, it was found that in just 30 generations, the best fitness result of each generation has risen by 68.4% with wheel sizes trending to a smaller size, fuel capacity to a smaller size, and cargo capacity to a larger size.
Co-Authors Arief Kelik Nugroho, Arief Kelik Prima Dewi Purnamasari Ratna, Anak Agung Putri