@article{ART003348248},
author={Fauzia Fika and Kim young in},
title={EEG-Based Classification of Pediatric ADHD via Multi-Classifier, Multi-Band Ablation: Verifying the Standalone Predictive Capacity of the Delta-Band Biomarker},
journal={ Journal of Software Forensics},
issn={3092-541X},
year={2026},
volume={22},
number={2},
pages={193-203},
doi={10.29056/jsf.2026.06.17}
TY - JOUR
AU - Fauzia Fika
AU - Kim young in
TI - EEG-Based Classification of Pediatric ADHD via Multi-Classifier, Multi-Band Ablation: Verifying the Standalone Predictive Capacity of the Delta-Band Biomarker
JO - Journal of Software Forensics
PY - 2026
VL - 22
IS - 2
PB - Korea Software Assessment and Valuation Society
SP - 193
EP - 203
SN - 3092-541X
AB - While EEG-based machine-learning models achieve high accuracy, purely data-driven feature selection may inadvertently obscure mutually correlated, biologically significant markers like the delta-band (0.5–3.0 Hz). To evaluate its independent predictive capacity, we propose a five-stage spatial–spectral ablation with four classifiers (Random Forest, SVM, XGBoost, and a Multilayer Perceptron). Using a public dataset of 61 ADHD and 60 control children (19 channels, 128 Hz), we extract 1,159 features and evaluate every feature-set under stratified 10-fold cross-validation. The full-feature model reaches 98.4–99.6% accuracy (0.998–1.000 AUC). Notably, utilizing only the 57 delta-band power/entropy features the models attained a mean AUC of 0.846, far exceeding a label-permuted control (0.499). These results provide strong evidence for the delta band as a standalone biomarker, highlighting the necessity of balancing mathematical optimization with biological interpretability in medical AI.
KW - Electroencephalography;Attention-deficit/hyperactivity disorder;Multi-classifier comparison;Ablation study;Feature-selection bias
DO - 10.29056/jsf.2026.06.17
ER -
Fauzia Fika and Kim young in. (2026). EEG-Based Classification of Pediatric ADHD via Multi-Classifier, Multi-Band Ablation: Verifying the Standalone Predictive Capacity of the Delta-Band Biomarker. Journal of Software Forensics, 22(2), 193-203.
Fauzia Fika and Kim young in. 2026, "EEG-Based Classification of Pediatric ADHD via Multi-Classifier, Multi-Band Ablation: Verifying the Standalone Predictive Capacity of the Delta-Band Biomarker", Journal of Software Forensics, vol.22, no.2 pp.193-203. Available from: doi:10.29056/jsf.2026.06.17
Fauzia Fika, Kim young in "EEG-Based Classification of Pediatric ADHD via Multi-Classifier, Multi-Band Ablation: Verifying the Standalone Predictive Capacity of the Delta-Band Biomarker" Journal of Software Forensics 22.2 pp.193-203 (2026) : 193.
Fauzia Fika, Kim young in. EEG-Based Classification of Pediatric ADHD via Multi-Classifier, Multi-Band Ablation: Verifying the Standalone Predictive Capacity of the Delta-Band Biomarker. 2026; 22(2), 193-203. Available from: doi:10.29056/jsf.2026.06.17
Fauzia Fika and Kim young in. "EEG-Based Classification of Pediatric ADHD via Multi-Classifier, Multi-Band Ablation: Verifying the Standalone Predictive Capacity of the Delta-Band Biomarker" Journal of Software Forensics 22, no.2 (2026) : 193-203.doi: 10.29056/jsf.2026.06.17
Fauzia Fika; Kim young in. EEG-Based Classification of Pediatric ADHD via Multi-Classifier, Multi-Band Ablation: Verifying the Standalone Predictive Capacity of the Delta-Band Biomarker. Journal of Software Forensics, 22(2), 193-203. doi: 10.29056/jsf.2026.06.17
Fauzia Fika; Kim young in. EEG-Based Classification of Pediatric ADHD via Multi-Classifier, Multi-Band Ablation: Verifying the Standalone Predictive Capacity of the Delta-Band Biomarker. Journal of Software Forensics. 2026; 22(2) 193-203. doi: 10.29056/jsf.2026.06.17
Fauzia Fika, Kim young in. EEG-Based Classification of Pediatric ADHD via Multi-Classifier, Multi-Band Ablation: Verifying the Standalone Predictive Capacity of the Delta-Band Biomarker. 2026; 22(2), 193-203. Available from: doi:10.29056/jsf.2026.06.17
Fauzia Fika and Kim young in. "EEG-Based Classification of Pediatric ADHD via Multi-Classifier, Multi-Band Ablation: Verifying the Standalone Predictive Capacity of the Delta-Band Biomarker" Journal of Software Forensics 22, no.2 (2026) : 193-203.doi: 10.29056/jsf.2026.06.17
