Thyroid Disease Using Hybrid ANN-RF-LR Framework with SHAP Explainability
International Journal on Advanced Science, Engineering and Information Technology · 12.10.2026 · 16 (5) · 1543-1549
DOI: 10.18517/ijaseit.16.5.14527 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal on Advanced Science, Engineering and Information Technology, online erschienen am 12.10.2026. Verfasst von Aqdas Bibi und 3 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Thyroid disease is a widespread endocrine disease that severely impacts the regulation of metabolic processes and general physiological systems and thus needs early and precise diagnosis to avoid serious complications. This paper presents a new hybrid prediction model that combines latent feature extraction using an Artificial Neural Network (ANN), a Random Forest (RF) classifier, and a Logistic Regression (LR) meta-learner to improve thyroid disease prediction. To address high-dimensional clinical data and class imbalance, the study uses a full preprocessing pipeline that includes categorical encoding, feature scaling, Principal Component Analysis (PCA), and Synthetic Minority Oversampling Technique (SMOTE). To test the proposed strategy, the study uses a large clinical dataset with 136,137 samples and 95 features. The experimental findings show that the hybrid ANN-RF-LR model performs better than traditional machine learning and isolated deep learning models, with higher accuracy, precision, and generalization. Moreover, SHapley Additive exPlanations (SHAP) is used to improve the interpretability of the model by determining the value of latent features used in decision-making. The findings indicate that the proposed framework effectively represents complex nonlinear relationships and provides transparency, supporting clinical trust. This research proposes a powerful, scalable decision-support platform to detect thyroid disease at its early stages and demonstrates the efficiency of deep representation learning with ensemble and meta-learning strategies in medical analytics.
Zitieren
APA
Bibi, A., Ahmed, I., Jo, J., & Yoon, Y. (2026). Thyroid Disease Using Hybrid ANN-RF-LR Framework with SHAP Explainability. International Journal on Advanced Science, Engineering and Information Technology, 16(5), 1543-1549. https://doi.org/10.18517/ijaseit.16.5.14527
Vancouver
Bibi A, Ahmed I, Jo J, Yoon Y. Thyroid Disease Using Hybrid ANN-RF-LR Framework with SHAP Explainability. International Journal on Advanced Science, Engineering and Information Technology. 2026;16(5):1543-1549. doi:10.18517/ijaseit.16.5.14527
BibTeX
@article{Bibi2026,
author = {Bibi, Aqdas and Ahmed, Iftikhar and Jo, Jaechoon and Yoon, Yeo-Chan},
title = {Thyroid Disease Using Hybrid ANN-RF-LR Framework with SHAP Explainability},
journal = {International Journal on Advanced Science, Engineering and Information Technology},
year = {2026},
volume = {16},
number = {5},
pages = {1543--1549},
doi = {10.18517/ijaseit.16.5.14527}
}
EUVIMED