Optimizing machine learning models for classification of stroke patients with epileptiform EEG pattern - the impact of dataset balancing techniques

Aug 4, 2026

Upload by Quinn

Epileptiform electroencephalogram (EEG) patterns are frequently observed in stroke patients and can meaningfully influence clinical management and outcomes. Building models to identify patients with a high probability of epileptiform EEG patterns may therefore support stroke care. Imbalanced datasets often bias machine learning toward dominant classes, reducing accuracy for underrepresented cases. This study compares models trained on imbalanced versus balanced data using four sampling strategies and ReliefF feature selection to quantify classification performance differences.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages10

Reading Time25 min

AccessFree to download

Abstract

Introduction

EEG patterns in stroke patients

Role of machine learning in classification tasks

[Available online at www.sciencedirect.com](Available online at www.sciencedirect.com) ScienceDirect Procedia Computer Science 246 (2024) 4600–4609 28th International Conference on Knowledge-Based and Intelligent Information & Engineering Systems (KES 2024) Optimizing machine learning models for classification of stroke patients with epileptiform EEG pattern: the impact of dataset balancing techniques Katerina Iscraa *, Alessandro Biscontina,b, Aleksandar Miladinovica,b, Andrea Boninia, Giovanni Furlanisc, Gabriele Prandinc, Michele Malesanic, Marcello Naccaratoc, Paolo Manganottic, Agostino Accardoa, Miloš Ajčevića aDepartement Engineering and Architecture, University of Trieste, Trieste, Italy bIstitute for Maternal and Child Health – IRCCS Burlo Garofolo, Trieste, Italy cClinical Unit of Neurology, Department of Medicine, Surgery and Health Sciences, Trieste University Hospital ASUGI, Trieste, Italy Abstract Epileptiform electroencephalogram (EEG) patterns are commonly observed in stroke patients and can significantly impact clinical management and patient outcomes. Therefore, the classification of the stroke patients in order to identify the subjects with high probability of epileptiform EEG patterns may improve the stroke management. In recent years, there has been a notable increase in interest and utilization of machine learning, especially in the domain of classification tasks. Nevertheless, the presence of imbalanced datasets presents hurdles for machine learning algorithms, resulting in skewed predictions toward dominant classes and diminished accuracy, especially for underrepresented ones. Hence, the study aims to evaluate the effects of dataset balancing methods on the classification efficacy of machine learning models for classification of stroke patients withepileptiform EEG patterns by conducting a comparative analysis between models trained on imbalanced and balanced datasets. Four different sampling techniques were employed: an oversampling technique, SMOTENC; an undersampling technique, NearMiss; and two techniques that combine over-and undersampling methods, SMOTEToken and SMOTEENN. The features selection was made using the ReliefF scoring method and for model construction, only features that presented a contribution value greater than 0.01 were utilized. Five different machine learning models were considered in the study: classification tree, logistic regression, naïve Bayes, artificial neural network and support vector machine. The produced models were trained on the original and resampled training set and subsequently the models' performances were evaluated on the test set. The results showed that SMOTENC was the most effective among the considered dataset balancing techniques, showing superior classification * Katerina Iscra. Tel.: +39 040 558 7130. E-mail address: katerina.iscra@phd.units.it 1877-0509 © 2024 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license ([https://creativecommons.org/licenses/by-nc-nd/4.0](https://creativecommons.org/licenses/by-nc-nd/4.0)) Peer-review under responsibility of the scientific committee of the 28th International Conference on Knowledge Based and Intelligent information and Engineering Systems 10.1016/j.procs.2024.09.324 Katerina Iscra et al. / Procedia Computer Science 246 (2024) 4600–4609 4601 performance compared to other methods and the original dataset. Models utilizing SMOTENC exhibited significant improvements in AUC (0.76 vs 0.67) and specificity values (0.73 vs 0.43) while maintaining comparable accuracy (0.72 vs 0.74) to those trained on the original dataset, respectively. Furthermore, it has been noted that different sampling techniques result indifferent selection of the most predictive features. In conclusion, our study highlights the crucial role of utilizing dataset balancing methods to improve the classification performances of predictive models in case of highly un

FAQs

Why do epileptiform EEG patterns matter in stroke care?

They are commonly observed in stroke patients and can significantly affect clinical management and patient outcomes. Accurate identification supports timely intervention and treatment planning.

What problem do imbalanced datasets create for machine learning classification?

Imbalanced datasets skew predictions toward dominant classes, which lowers accuracy—especially for underrepresented groups. This reduces the reliability of predictive models.

Which dataset balancing technique performed best and how was it evaluated?

SMOTENC was the most effective among the tested techniques. Models using SMOTENC achieved higher AUC and specificity while maintaining comparable accuracy compared with models trained on the original dataset.

Aug 4, 2026

Upload by Quinn