Article Per Year (5 Year)
p-Index From 2021 - 2026
0.408
P-Index
This Author published in this journals
All Journal Kinetik: Game Technology, Information System, Computer Network, Computing, Electronics, and Control Indonesian Journal of Data and Science
Azlan Mohd Zain
Universiti Teknologi Malaysia
Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management Electrical & Electronics Engineering Energy Engineering Mathematics

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

Found 2 Documents
Search

 1 
Comparative Evaluation of Machine Learning Models for Heavy Crude Oil Viscosity Prediction Using Repeated Nested Cross-Validation and Independent Holdout Testing Enggie Hendrawan Saputra; Ilham Ari Elbaith Zaeni; Didik Dwi Prasetya; Azlan Mohd Zain; Welly Antonius; I Made Wirawan
Indonesian Journal of Data and Science Vol. 7 No. 2 (2026): Indonesian Journal of Data and Science
Publisher : yocto brain

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.56705/ijodas.v7i2.455

Abstract

Accurate prediction of heavy crude oil viscosity supports reservoir engineering, production planning, and flow assurance. This study presents a rigorous comparative evaluation rather than a new machine-learning framework. The published dataset, derived from Kamel et al. and reproduced by Li et al., represents 28 Middle Eastern heavy crude-oil samples through 196 development measurements and 47 independent holdout measurements at 20–80 °C. The supplied workbook contains API gravity, temperature, C1, C2, C3, C4–C6, C7+, and viscosity, with a development viscosity range of 632.88–1267.65 cP; however, it does not include row-level oil or reservoir identifiers. Linear Regression, the Beggs–Robinson correlation, SVR, Random Forest, and Gradient Boosting were evaluated. Imputation, scaling, and hyperparameter selection were embedded in repeated nested cross-validation with five outer folds repeated twice and five inner folds. Final models were evaluated once on the untouched 47-record holdout, with bootstrap confidence intervals, corrected pairwise tests, residual diagnostics, and held-out permutation importance. Gradient Boosting achieved an internal R² of 0.99313 (95% CI: 0.99077–0.99549) and RMSE of 11.41 cP (10.03–12.80). On the independent holdout, it achieved R² = 0.99308 (bootstrap 95% CI: 0.98772–0.99637), RMSE = 8.43 cP, MAE = 6.64 cP, and MAPE = 0.78%. Corrected comparisons showed lower RMSE than Random Forest and SVR. Holdout diagnostics detected no statistically significant heteroscedasticity for Gradient Boosting, and extreme-value sensitivity produced similar performance. Temperature and C7+ were the dominant held-out predictors. These results are encouraging within the sampled domain, but the absent oil identifiers prevent group-disjoint validation, and no external reservoir dataset supports field-level generalization.
Comparative Evaluation of BM25–FAISS and Small-LLM–GPT in Retrieval-Augmented Generation Concept Map Assessment Maskur Maskur; Didik Dwi Prasetya; Triyanna Widiyaningtyas; Azlan Mohd Zain
Kinetik: Game Technology, Information System, Computer Network, Computing, Electronics, and Control Vol. 11, No. 1, February 2026
Publisher : Universitas Muhammadiyah Malang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.22219/kinetik.v11i1.2594

Abstract

Concept map-based assessment is a practical approach to measure students’ conceptual understanding, but manual assessment still faces challenges such as subjectivity, inconsistency, and limited scalability. This study proposes the application of Retrieval-Augmented Generation (RAG) as an artificial intelligence-based automated assessment solution in an educational context. The objectives of this study are to compare the effectiveness of two retrieval methods, BM25 and FAISS, and to analyse the trade-off between large-scale generative models (GPT) and Small-LLM in assessing concept map propositions. This study uses a quantitative experimental approach by combining a retriever and a generator in the RAG system. Performance evaluation is carried out using the Macro-F1 and QWK metrics to measure agreement with expert judgment, and the Explanation Relevance Score (ERS) to assess explanation quality. The experimental results show that the FAISS–GPT combination achieves the best performance, with a Macro-F1 of 0.338 and a QWK of 0.146, slightly superior to the BM25–GPT combination. In contrast, the use of Small-LLM, both with BM25 and FAISS, showed lower performance with Macro-F1 values in the range of 0.167–0.221 and QWK close to zero. This finding confirms that semantic-based retrieval plays a vital role in improving the accuracy of automated assessment, while large-scale generative models are more effective in representing conceptual relationships in depth. This study contributes through a comparative analysis of retrievers and generators, and by introducing ERS as an additional metric for RAG-based automated assessment in the field of education.
Co-Authors Didik Dwi Prasetya Enggie Hendrawan Saputra I Made Wirawan Maskur Maskur Triyanna Widiyaningtyas Welly Antonius Zaeni, Ilham Ari Elbaith