Evaluating EEG-Based Seizure Classification Using Foundation and Classical Ensemble Models
Source abstract
Abstract
Electroencephalogram (EEG)-based seizure classification remains challenging due to inter-subject variability and heterogeneous signal characteristics. Foundation models offer a promising alternative to dataset-specific training by leveraging pretrained priors. In this study, we evaluate a tabular foundation model, the Tabular Prior-Data Fitted Network (TabPFN), against classical ensemble baselines (gradient boosting, random forests, AdaBoost, and XGBoost) for EEG seizure segment classification. We use subject-independent GroupKFold cross-validation without out-of-fold evaluation to assess generalization to unseen individuals. Experiments on the Bangalore EEG Epilepsy Dataset (BEED) and the University of Bonn (Bonn) dataset show that TabPFN achieves higher accuracy than classical ensembles, reaching 99.7% on BEED and 99.6% on Bonn. These results suggest that pretrained tabular priors can be effective in feature-based EEG pipelines where subject-level generalization is required.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: EEG and Brain-Computer Interfaces · Epilepsy research and treatment · Functional Brain Connectivity Studies
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
Center for Artificial Intelligence and Multidisciplinary Innovations, Department of Auditing, College of Accounting Sciences, University of South Africa, Pretoria 0002, South Africa
References (61)
Show source-provided references
1. EEG based multi-class seizure type classification using convolutional neural network and transfer learning10.1016/j.neunet.2020.01.017 ↗
2. Channel reflection: Knowledge-driven data augmentation for EEG-based brain–computer interfaces10.1016/j.neunet.2024.106351 ↗
3. Automated multi-class seizure-type classification system using EEG signals and machine learning algorithms10.1109/ACCESS.2024.3462772 ↗
4. Electroencephalogram (EEG) based fuzzy logic and spiking neural networks (FLSNN) for advanced multiple neurological disorder diagnosis10.1007/s10548-025-01106-1 ↗
5. Harnessing electroencephalography connectomes for cognitive and clinical neuroscience10.1038/s41551-025-01442-4 ↗
6. The applied principles of EEG analysis methods in neuroscience and clinical neurology
7. A Review of Machine Learning and Deep Learning Trends in EEG-Based Epileptic Seizure Prediction10.1109/ACCESS.2025.3606966 ↗
8. Zhang, X., Zhang, X., Huang, Q., and Chen, F. (2024). A Review of Epilepsy Detection and Prediction Methods Based on EEG Signal Processing and Deep Learning. Front. Neurosci., 18.10.3389/fnins.2024.1468967 ↗
9. Evaluating the clinical readiness of artificial intelligence in EEG-Based epilepsy diagnosis
10. Hu, Y., Liu, J., Sun, R., Yu, Y., and Sui, Y. (2024). Classification of Epileptic Seizures in EEG Data Based on Iterative Gated Graph Convolution Network. Front. Comput. Neurosci., 18.10.3389/fncom.2024.1454529 ↗
11. Tu, S., Cao, L., Zhang, D., Chen, J., Ma, L., Zhang, Y., and Yang, Y. (2024, January 9–15). DMNet: Self-Comparison Driven Model for Subject-Independent Seizure Detection. Proceedings of the Advances in Neural Information Processing Systems (NeurIPS), Vancouver, BC, Canada.
12. D’Amora, M., Galgani, A., Marchese, M., Tantussi, F., Faraguna, U., De Angelis, F., and Giorgi, F.S. (2023). Zebrafish as an innovative tool for epilepsy modeling: State of the art and potential future directions. Int. J. Mol. Sci., 24.10.3390/ijms24097702 ↗
13. Non-electroencephalogram-based seizure detection devices: State of the art and future perspectives10.1016/j.yebeh.2023.109486 ↗
14. Manole, A.M., Sirbu, C.A., Mititelu, M.R., Vasiliu, O., Lorusso, L., Sirbu, O.M., and Ionita Radu, F. (2023). State of the art and challenges in epilepsy—A narrative review. J. Pers. Med., 13.10.3390/jpm13040623 ↗
15. Espinosa, N., Amorim, A., and Huebner, R. (2020, January 27–28). Feedforward Neural Network with Backpropagation for Epilepsy Seizure Detection. Proceedings of the IEEE EMBS International StudentConference 2020 in Latin America, Quito, Ecuador.
16. Time-frequency domain deep convolutional neural network for the classification of focal and non-focal EEG signals10.1109/JSEN.2019.2956072 ↗
17. Classification of electroencephalogram signals with combined time and frequency features10.1016/j.eswa.2011.02.110 ↗
18. Electroencephalogram (EEG) time series classification: Applications in epilepsy10.1007/s10479-006-0076-x ↗
19. Epileptic seizure detection in EEGs using time–frequency analysis10.1109/TITB.2009.2017939 ↗
20. Investigation of epileptic seizure signatures classification in EEG using supervised machine learning algorithms10.18280/ts.400104 ↗
21. Accurate Predictions on Small Data with a Tabular Foundation Model10.1038/s41586-024-08328-6 ↗
22. Ruan, Y., Lan, X., Ma, J., Dong, Y., He, K., and Feng, M. (2024). Language modeling on tabular data: A survey of foundations, techniques and evolution. arXiv.
23. Revisiting deep learning models for tabular data
24. Huang, X., Khetan, A., Cvitkovic, M., and Karnin, Z. (2020). Tabtransformer: Tabular data modeling using contextual embeddings. arXiv.
25. Deep neural networks and tabular data: A survey10.1109/TNNLS.2022.3229161 ↗
26. EEG signal classification using PCA, ICA, LDA and support vector machines10.1016/j.eswa.2010.06.065 ↗
27. Kapoor, B., Nagpal, B., Jain, P.K., Abraham, A., and Gabralla, L.A. (2022). Epileptic seizure prediction based on hybrid seek optimization tuned ensemble classifier using EEG signals. Sensors, 23.10.3390/s23010423 ↗
28. Hollmann, N., Müller, S., Eggensperger, K., and Hutter, F. (2022). Tabpfn: A transformer that solves small tabular classification problems in a second. arXiv.
29. Epileptic seizure detection using dynamic wavelet network10.1016/j.eswa.2005.04.007 ↗
30. Bose, S., Rama, V., and Rao, C.R. (2017). EEG signal analysis for seizure detection using discrete wavelet transform and random forest. 2017 International Conference on Computer and Applications (ICCA), IEEE.10.1109/COMAPP.2017.8079760 ↗
31. Bhattacharyya, S., Konar, A., Tibarewala, D., Khasnobish, A., and Janarthanan, R. (2014). Performance analysis of ensemble methods for multi-class classification of motor imagery EEG signal. 2014 International Conference on Control, Instrumentation, Energy and Communication (CIEC), IEEE.10.1109/CIEC.2014.6959183 ↗
32. EEG based automated detection of seizure using machine learning approach and traditional features10.1016/j.eswa.2024.123991 ↗
33. Paneru, B. (2025). Interpretable Machine Learning for Epileptic Seizure Detection on the BEED Using LIME with an Ensemble Network. medRxiv.10.1101/2025.09.30.25336996 ↗
34. Banu, P.K.N., Azar, A.T., and Kamal, N.A. (2024). Impact of Multi-domain Features for EEG Based Epileptic Seizures Classification. International Conference on Advanced Intelligent Systems and Informatics, Springer.
35. Alshaya, H., and Hussain, M. (2023). EEG-based classification of epileptic seizure types using deep network model. Mathematics, 11.10.3390/math11102286 ↗
36. A long short-term memory deep learning network for the prediction of epileptic seizures using EEG signals10.1016/j.compbiomed.2018.05.019 ↗
37. Efficient epileptic seizure prediction based on deep learning10.1109/TBCAS.2019.2929053 ↗
38. Mallick, S., and Baths, V. (2024). Novel deep learning framework for detection of epileptic seizures using EEG signals. Front. Comput. Neurosci., 18.10.3389/fncom.2024.1340251 ↗
39. Feature Engineering for Epileptic Seizure Classification Using SeqBoostNet
40. Sajid, M.S., Ghosh, A.K., and Nusrat, F. (2025). PaperNet: Efficient Temporal Convolutions and Channel Residual Attention for EEG Epilepsy Detection. arXiv.
41. Ye, H.J., Liu, S.Y., and Chao, W.L. (2025). A closer look at tabpfn v2: Strength, limitation, and extension. arXiv.
42. Grinsztajn, L., Flöge, K., Key, O., Birkel, F., Jund, P., Roof, B., Jäger, B., Safaric, D., Alessi, S., and Hayler, A. (2025). TabPFN-2.5: Advancing the State of the Art in Tabular Foundation Models. arXiv.
43. and Banu, P.K.N. (2025, June 28). BEED: Bangalore EEG Epilepsy Dataset. UCI Machine Learning Repository. Available online: https://archive.ics.uci.edu/dataset/1134/beed:+bangalore+eeg+epilepsy+dataset.
44. An epileptic EEG detection method based on data augmentation and lightweight neural network10.1109/JTEHM.2023.3308196 ↗
45. Indications of nonlinear deterministic and finite-dimensional structures in time series of brain electrical activity: Dependence on recording region and brain state10.1103/PhysRevE.64.061907 ↗
46. Parmar, A., Katariya, R., and Patel, V. (2018). A review on random forest: An ensemble classifier. International Conference on Intelligent Data Communication Technologies and Internet of Things, Springer.10.1007/978-3-030-03146-6_86 ↗
47. Mienye, I.D., Obaido, G., Aruleba, K., and Dada, O.A. (2021). Enhanced prediction of chronic kidney disease using feature selection and boosted classifiers. International Conference on Intelligent Systems Design and Applications, Springer.10.1007/978-3-030-96308-8_49 ↗
48. Natekin, A., and Knoll, A. (2013). Gradient boosting machines, a tutorial. Front. Neurorobot., 7.10.3389/fnbot.2013.00021 ↗
49. Mienye, I.D., and Jere, N. (2024). Optimized ensemble learning approach with explainable AI for improved heart disease prediction. Information, 15.10.3390/info15070394 ↗
50. Chen, T., He, T., Benesty, M., Khotilovich, V., Tang, Y., Cho, H., Chen, K., Mitchell, R., Cano, I., and Zhou, T. (2015). Xgboost: Extreme Gradient Boosting. Available online: http://r.meteo.uni.wroc.pl/web/packages/xgboost/vignettes/xgboost.pdf.10.32614/CRAN.package.xgboost ↗
51. Schapire, R.E. (2013). Explaining adaboost. Empirical Inference: Festschrift in Honor of Vladimir N. Vapnik, Springer.10.1007/978-3-642-41136-6_5 ↗
52. Yin, M., Wortman Vaughan, J., and Wallach, H. (2019, January 4–9). Understanding the effect of accuracy on trust in machine learning models. Proceedings of the 2019 Chi Conference on Human Factors in Computing Systems, Glasgow, UK.10.1145/3290605.3300509 ↗
53. Amrouche, S., Basara, L., Calafiura, P., Estrade, V., Farrell, S., Ferreira, D.R., Finnie, L., Finnie, N., Germain, C., and Gligorov, V.V. (2019). The tracking machine learning challenge: Accuracy phase. The NeurIPS’18 Competition: From Machine Learning to Intelligent Conversations, Springer.10.1007/978-3-030-29135-8_9 ↗
54. Deep learning for electroencephalogram (EEG) classification tasks: A review10.1088/1741-2552/ab0ab5 ↗
55. Association of recall rates with sensitivity and positive predictive values of screening mammography10.2214/ajr.177.3.1770543 ↗
56. Chicco, D., and Jurman, G. (2020). The advantages of the Matthews correlation coefficient (MCC) over F1 score and accuracy in binary classification evaluation. BMC Genom., 21.10.1186/s12864-019-6413-7 ↗
57. Obaido, G., Ogbuokiri, B., Mienye, I.D., and Kasongo, S.M. (2022). A voting classifier for mortality prediction post-thoracic surgery. International Conference on Intelligent Systems Design and Applications, Springer.10.1007/978-3-031-35501-1_26 ↗
58. Obaido, G., Ogbuokiri, B., Swart, T.G., Ayawei, N., Kasongo, S.M., Aruleba, K., Mienye, I.D., Aruleba, I., Chukwu, W., and Osaye, F. (2022). An interpretable machine learning approach for hepatitis b diagnosis. Appl. Sci., 12.10.3390/app122111127 ↗
59. Gorriz, J.M., Clemente, R.M., Segovia, F., Ramirez, J., Ortiz, A., and Suckling, J. (2024). Is K-fold cross validation the best model selection method for Machine Learning?. arXiv.
60. Kumar, D.S., Sriram, S., and Kumar, P. (2025). Enhancing Greenhouse Farming Through Machine Learning: A Focus on Feature Importance. 2025 International Conference on Computing Technologies (ICOCT), IEEE.10.1109/ICOCT64433.2025.11118524 ↗
61. Akbari, H., Esmaili, S.S., and Zadeh, S.F. (2019). Classification of seizure and seizure-free EEG signals based on empirical wavelet transform and phase space reconstruction. arXiv.