Hybrid Ensemble Learning for EEG-Based Epileptic Seizure Forecasting

Mason Dana; Khandaker Mamun Ahmed

cs.LGSep 26, 2026

Hybrid Ensemble Learning for EEG-Based Epileptic Seizure Forecasting

Authors: Mason Dana, Khandaker Mamun Ahmed

Organizations: The Beacom College of Computer and Cyber Sciences, Dakota State University Madison, SD, USA

Abstract

Epileptic seizure forecasting aims to provide actionable warnings before seizure onset, yet patient-independent generalization and false-alarm control remain major challenges. We propose a calibrated hybrid ensemble for EEG-based seizure forecasting that combines five deep learning models and three classical machine learning models through a logistic regression stacking meta-learner. The proposed pipeline integrates signal preprocessing, handcrafted feature extraction, class-imbalance handling, probability calibration, and clinically motivated post-processing. We evaluate the framework on CHB-MIT using strict Leave-One-Patient-Out (LOPO) cross-validation, with threshold and post-processing parameters selected only on held-out meta data. On the filtered cohort, excluding patients with anomalous preictal rates below 1% or above 15%, the model achieves 74.2% seizure-level sensitivity at 1.24 false alarms per hour, with an average warning time of 16.9 minutes. A test-tuned oracle constrained to the target false-alarm budget achieves 60.9% sensitivity at 0.951 false alarms per hour, highlighting the importance of reporting sensitivity together with realized false-alarm rates. Our code is available at: https://github.com/DanaMason/IEEE-CARS-Hybrid-Ensemble-Learning-for-EEG-Based-Epileptic-Seizure-Forecasting

Figures & tables

Fig. 1 : Overview of the proposed seizure-forecasting framework, including data acquisition, preprocessing, feature extraction, LOPO-based partitioning, deep learning and classical model training, stacking-based meta-learning, temporal post-processing, and performance evaluation. Signal preprocessing is applied only to the classical machine learning branch.

Patient	Sens.	Detected	FA/hr	Warn (min)
chb01	28.6%	2/7	1.16	15.5
chb02	100.0%	2/2	1.24	10.3
chb03	28.6%	2/7	0.52	9.6
chb04	100.0%	4/4	1.46	12.7
chb05	60.0%	3/5	0.74	18.3
chb06	50.0%	5/10	0.57	13.7

TABLE I : Per-patient seizure-level performance under LOPO cross-validation using a 30-minute preictal horizon and 5-minute exclusion gap. † denotes patients with anomalous preictal rates excluded from the filtered aggregate.

Fig. 2 : Sensitivity and FA/hr for meta-tuned and test-tuned configurations under fixed post-processing settings.

Metric	All ( $n{=}24$ )	Filtered ( $n{=}20$ )
Sensitivity	$0.001\pm 0.000$	$0.001\pm 0.000$
Specificity	$0.999\pm 0.000$	$0.999\pm 0.000$
Precision	$0.104\pm 0.077$	$0.086\pm 0.046$
F1-score	$0.002\pm 0.001$	$0.002\pm 0.001$
Accuracy	$0.890\pm 0.092$	$0.925\pm 0.040$
AUC-ROC	$0.487\pm 0.085$	$0.485\pm 0.091$

TABLE II : Average window-level performance, reported as mean

\pm

standard deviation for all and filtered patients.

Method	Year	Sens.	FA/hr	Threshold
Patient-specific evaluation
Truong et al. [ 20 ]	2018	0.812	0.159	N/R
Ben Messaoud & Chavez [ 25 ]	2021	0.821	0.080	N/R
Patient-independent LOPO evaluation
Meng et al. (3D-SERESNet) [ 26 ]	2025	0.844	0.232	N/R
Ours, honest (all, $n{=}24$ )	2026	0.728	1.251	Meta slice

TABLE III : Comparison of CHB-MIT seizure forecasting results across evaluation protocols 2 2 2 The threshold-source column indicates whether the reported operating point was selected using a held-out meta slice, test labels, or was not reported. Results across protocol groups are not directly comparable. .

Ensemble Learning

Ensemble learning combines multiple predictors to improve accuracy, robustness, calibration, or uncertainty estimates.

See all 272 papers

EEG-FuseFormer: A Transformer-Driven Feature Fusion Framework for Seizure Onset Prediction
Jun 1, 2026Vigneshwar Hariharan, Chithra Reghuvaran, Arlene John +4Electroencephalography EEG Seizure Detection
National University of Singapore · University College Dublin · University of Twente +1
DiffEEG: A Self-Supervised Denoising Diffusion Model for Learning EEG seizure-related Representations
Jul 13, 2026Abdulkader Helwan, Lina Abou-Abbas, Hussein El Amouri +2Electroencephalography EEG Seizure Detection
Department of Electrical and Computer Engineering, Lebanese American University, Byblos, Lebanon · Institute of Applied Artificial Intelligence, TÉLUQ University, Montreal, Canada
Epileptic Seizure Prediction Using Patient-Adaptive Transformer Networks
Mar 26, 2026Mohamed Mahdi, Asma BaghdadiEEG Seizure Detection Electroencephalography
Information and Communication Engineering Department National Engineering School of Tunis Tunis El Manar University · REGIM-Lab, University of Sfax, National Engineering School of Sfax (ENIS), BP 1173, Sfax, 3038, Tunisia

Jun 1, 2026J/K move · Enter open · S save

EEG-FuseFormer: A Transformer-Driven Feature Fusion Framework for Seizure Onset Prediction

Vigneshwar Hariharan, Chithra Reghuvaran, Arlene John, Nhat Pham+3

National University of Singapore · University College Dublin · University of Twente+1

Electroencephalography EEG Seizure Detection Hybrid CNN-Transformer Architectures Feature Fusion

Open paper AskPDF arXiv

Epilepsy is one of the most common neurological disorders globally, characterized by recurring seizures and significantly impacting the quality of life. Despite advancements in diagnostic techniques, the mitigation of risks faced by epilepsy patients remains challenging due to the unpredictability of seizure events. An accurate forecast of seizure onset helps to reduce risks in epilepsy patients. In this paper, we propose EEG-FuseFormer, a transformer-based feature fusion framework for seizure-onset prediction that combines intermediate features extracted from Convolutional Neural Networks-Long Short-Term Memory (CNN-LSTM) and ResNet-18 networks. The CNN-LSTM architecture captures both spatial and temporal features directly from the raw signal, whereas the ResNet-18 extracts features from the Short-Time Fourier Transform (STFT) representation of the EEG signals. Fusion is carried out using a transformer encoder, and the final prediction is generated using fully connected dense layers. The CHB-MIT dataset was used to validate the proposed model. The results show that the proposed model achieves a mean recall of 98.85% and outperforms most of the state-of-the-art methods. This study evaluates the ability of the proposed feature fusion model to generalize in cross-patient testing scenarios. Fine-tuning pre-trained models on limited target patient data (target adaptation) within the cross-patient validation framework results in higher recall, precision, and F1-score metrics in comparison to the conventional cross-patient validation approach. Finally, the runtime-based computational complexity of the model is assessed across diverse hardware platforms to highlight the performance-complexity trade-off.

Abstract

Ensemble Learning

Explore similar work

EEG-FuseFormer: A Transformer-Driven Feature Fusion Framework for Seizure Onset Prediction

DiffEEG: A Self-Supervised Denoising Diffusion Model for Learning EEG seizure-related Representations

Epileptic Seizure Prediction Using Patient-Adaptive Transformer Networks