Article Per Year (5 Year)
p-Index From 2021 - 2026
0.408
P-Index
This Author published in this journals
All Journal Journal of Student Research Exploration Journal of Soft Computing Exploration
Sidiq Noor Kharisma
Department of Computer Science, Universitas Negeri Semarang, Indonesia
Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management Electrical & Electronics Engineering

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

Found 2 Documents
Search

 1 
Enhanced Out-of-Fold Stacking with Feature Grouping and Model-Specific Transformations for Diabetes Prediction Improvement Ari Nugroho Putro; Sidiq Noor Kharisma; Gea Destadia Al-Zahra; Much Aziz Muslim; Dwika Ananda Agustina Pertiwi
Journal of Student Research Exploration Vol. 4 No. 1 (2026): January 2026
Publisher : SHM Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.52465/josre.v4i1.674

Abstract

Diabetes mellitus is a chronic disease with serious implications for global health. Early detection is essential to reduce these risks, and machine learning methods are widely used in diabetes prediction. However, improving accuracy remains a major challenge in the development of predictive models. This study proposes a stacking-based ensemble learning approach with an out-of-fold (OOF) scheme to improve classification performance. The proposed method consists of several systematic steps, namely (1) data preprocessing via median imputation of invalid values and feature transformation according to model characteristics, (2) the creation of base learners comprising Logistic Regression, Gaussian Naïve Bayes, Support Vector Machine, Random Forest, and XGBoost, (3) model training using Stratified Cross Validation 5 Fold to generate OOF predictions, (4) combining all OOF predictions into a meta-feature matrix, and (5) training an XGBoost-based meta-model to generate the final prediction. This approach enables the meta-model to optimally learn the relationships among the outputs of the baseline models. Experimental results show that the proposed method achieves an accuracy of 91.15%, precision of 90.65%, recall of 83.21%, and an F1-score of 86.77%. These results indicate that stacking is effective in improving the accuracy of diabetes predictions.
Comparative performance of MLP, CNN, TabNet, and FT-Transformer for credit risk prediction with TPOT  hyperparameter optimization Much Aziz Muslim; Dwika Ananda Agustina Pertiwi; Jumanto Unjung; Sidiq Noor Kharisma; Gea Destadia Al-Zahra; Shahrul Nizam Bin Salahudin
Journal of Soft Computing Exploration Vol. 7 No. 3 (2026): September 2026
Publisher : SHM Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.52465/joscex.v7i3.250

Abstract

Credit risk prediction is important for supporting accurate lending decisions and reducing the potential risk of loan default. However, deep learning performance on tabular credit data can vary according to model architecture and TPOT hyperparameter configuration. Therefore, this study aims to comparatively evaluate Multilayer Perceptron, Convolutional Neural Network, TabNet, and FT-Transformer for credit risk prediction and examine the effect of TPOT hyperparameter optimization. The study uses the Lending Club dataset from 2007–2020, with 2,000 records from 2019–2020 sampled for analysis. The data were preprocessed through feature selection, missing-value handling, outlier treatment, categorical encoding, and feature scaling, while Synthetic Minority Over-sampling Technique was applied to address class imbalance. A stratified 70:30 train-test split was used, and the models were evaluated using accuracy, precision, recall, and F1-score before and after optimization. The results show that all four models improved after optimization, although the magnitude varied. Convolutional Neural Network achieved the highest optimized accuracy of 95.50%, while TabNet achieved the highest recall of 86.05% and F1-score of 86.29%. Multilayer Perceptron obtained the highest precision of 91.02%, whereas TabNet achieved the largest F1-score improvement of 6.84 percentage points. Overall, TPOT hyperparameter optimization improved model performance, with TabNet providing the most favorable overall balance for credit risk prediction.
Co-Authors Ari Nugroho Putro Dwika Ananda Agustina Pertiwi Dwika Ananda Agustina Pertiwi Gea Destadia Al-Zahra Jumanto Unjung Much Aziz Muslim Shahrul Nizam Bin Salahudin