SavedAdd your journal

Journals / Hacettepe Journal of Mathematics and Statistics / 2026 / Volume: 55 Issue: 3

HypoGAN: A mode- and boundary-aware generative oversampling framework for imbalanced tabular data

Olcay Alpay(SINOP UNIVERSITY)0000-0003-1446-0801
Pages
—
Publication Date
2026-06-30
🏛️Bibliographic metadata and full text access for this article are provided via DergiPark.
DergiPark Record ↗
✨

AI Summary & Key Takeaways

AI Insight
⚡ TL;DR(Too Long; Didn't Read)

The results indicate that HypoGAN is a competitive oversampling framework in a range of challenging imbalance settings.

📌 3 Key Takeaways

🎯
Objective & Hypothesis:While numerous oversampling strategies have emerged in recent years, synthesizing minority-class samples that preserve the intrinsic data structure while improving classifier performance remains challenging, especially under severe class imbalance.
📊
Key Findings & Data:In this study, we propose HypoGAN, a generative oversampling framework for imbalanced numeric tabular data that integrates minority mode discovery, local minority-majority boundary information, latent-noise-driven candidate generation, and post-generation filtering to produce classification-useful minority samples.
💡
Conclusion & Impact:The results indicate that HypoGAN is a competitive oversampling framework in a range of challenging imbalance settings.
🔬Design:Research Article
📍Scope:Comprehensive Sample
⚙️Variables:Generative adversarial networks · HypoGAN · imbalanced learning

Abstract

While numerous oversampling strategies have emerged in recent years, synthesizing minority-class samples that preserve the intrinsic data structure while improving classifier performance remains challenging, especially under severe class imbalance. In this study, we propose HypoGAN, a generative oversampling framework for imbalanced numeric tabular data that integrates minority mode discovery, local minority-majority boundary information, latent-noise-driven candidate generation, and post-generation filtering to produce classification-useful minority samples. To evaluate the proposed framework, we adopted a two-phase experimental strategy: (i) simulation experiments under class imbalance ratios of 90:10, 95:5, and 98:2 with feature dimensionalities of $n = 3, 5, 10, 20$, and (ii) real-world experiments on Wisconsin Breast Cancer, Pima Indians Diabetes, and detection of credit card fraud. In all experiments, HypoGAN was compared with SMOTE, ADASYN, and Borderline-SMOTE within a nested train/validation/test framework with leakage protection. The results indicate that HypoGAN is a competitive oversampling framework in a range of challenging imbalance settings. In the simulation study, it achieved particularly strong performance in lower- and medium-dimensional scenarios, while remaining precision-oriented under more difficult conditions. In real-world experiments, HypoGAN remained competitive in the Wisconsin Breast Cancer and Credit Card Fraud Detection datasets, achieving F1-scores of 0.9489 and 0.9028, respectively, compared to 0.9489 and 0.9037 for SMOTE. Additional results suggest that HypoGAN’s performance is scenariodependent, with effectiveness influenced by the dataset’s structure and the characteristics of the imbalance.

🔖

Cite This Article

Alpay, O. (2026). HypoGAN: A mode- and boundary-aware generative oversampling framework for imbalanced tabular data. Hacettepe Journal of Mathematics and Statistics, Volume: 55 Issue: 3. https://doi.org/10.15672/hujms.1902068

📄 Full Text PDF Viewer

📄

PDF görüntüleyici hazırlanıyor...

References (28)

  1. N.V. Chawla, K.W. Bowyer, L.O. Hall and W.P. Kegelmeyer, SMOTE: Synthetic minority over-sampling technique, J. Artif. Intell. Res. 16, 321-357, 2002.
  2. Y. Chen, R. Chang and J. Guo, Effects of data augmentation method borderline- SMOTE on emotion recognition of EEG signals based on convolutional neural network, IEEE Access 9, 47491-47502, 2021.
  3. I. Dey and V. Pratap, A comparative study of SMOTE, borderline-SMOTE, and ADASYN oversampling techniques using different classifiers, Proc. 3rd Int. Conf. Smart Data Intell., 294-302, 2023.
  4. D. Dua and C. Graff, UCI Machine Learning Repository, University of California, Irvine, School of Information and Computer Sciences, 2019. [Online]. Available: http: //archive.ics.uci.edu/ml. [Accessed: 27-Apr-2026].
  5. G. Eom and H. Byeon, Searching for optimal oversampling to process imbalanced data: Generative adversarial networks and synthetic minority over-sampling technique, Mathematics 11, 3605, 2023.
  6. A. Gosain and S. Sardana, Handling class unbalance problem using oversampling techniques: A review, Proc. Int. Conf. Adv. Comput. Commun. Informat., 79-85, 2017.
  7. H. Han, W.Y.Wang and B.H. Mao, Borderline-SMOTE: A new over-sampling method in imbalanced data sets learning, in: Intelligent computing, Springer, Berlin, 878-887, 2005.
  8. H. He, Y. Bai, E.A. Garcia and S. Li, ADASYN: Adaptive synthetic sampling approach for unbalanced learning, Proc. IEEE Int. Joint Conf. Neural Netw., 1322-1328, 2008.
  9. T.R. Hoens and N.V. Chawla, Imbalanced datasets: From sampling to classifiers, in: H. He and Y. Ma (eds), Imbalanced learning: Foundations, algorithms, and applications, Wiley, 43-59, 2013.
  10. L. Hussain, K.J. Lone, I.A. Awan, A.A. Abbasi and J.U.R. Pirzada, Detecting congestive heart failure by extracting multimodal features with SMOTE for imbalanced data using robust ML techniques, Waves Random Complex Media 32, 1079-1102, 2022.
  11. C. Jiang, W. Lv and J. Li, Protein-protein interaction sites prediction using batch normalization based CNNs and borderline-SMOTE, IEEE/ACM Trans. Comput. Biol. Bioinf. 20, 2190-2199, 2023.
  12. Q. Jin, R. Lin and F. Yang, E-WACGAN: Enhanced generative model of signaling data based on WGAN-GP and ACGAN, IEEE Syst. J. 14, 3289-3300, 2019.
  13. T. Kari, L. Du, A. Rouzi, X. Ma, Z. Liu and B. Li, Fault diagnosis of power transformer based on improved ACGAN under imbalanced data, Comput. Mater. Continua 75 (2), 4573, 2023.
  14. W. Lee, B. Noh, Y. Kim and K. Jeong, Generation of network traffic using WGAN-GP and a DFT filter for resolving data imbalance, in: Internet and distributed computing systems, Springer, 2019.
  15. A.F.N. Masruriyah, H.Y. Novita, C.E. Sukmawati, A. Fauzi, D. Wahiddin and H.H. Handayani, Thorough evaluation of the effectiveness of SMOTE and ADASYN oversampling methods in enhancing supervised learning performance for imbalanced heart disease datasets, in: Proc. 8th Int. Conf. Informatics and Computing (ICIC), 1-7, IEEE, 2023.
  16. M. Mirza and S. Osindero, Conditional generative adversarial nets, arXiv:1411.1784 [cs.LG], 2014.
  17. E. Nazari, P. Branco and G.V. Jourdan, Using CGAN to address class imbalance and small sample size in cybersecurity, Proc. 18th Int. Conf. Privacy Security Trust, 1-10, 2021.
  18. A. Odena, C. Olah and J. Shlens, Conditional image synthesis with auxiliary classifier GANs, Proc. 34th Int. Conf. Mach. Learn., 2642-2651, 2017.
  19. G.A. Pradipta, R. Wardoyo, A. Musdholifah, I.N.H. Sanjaya and M. Ismail, SMOTE for handling imbalanced data problem: A review, in: Proc. 6th Int. Conf. Informatics and Computing (ICIC), 1-8, IEEE, 2021.
  20. F.R.A. Pratama and S.I. Oktora, Synthetic minority over-sampling technique (SMOTE) for handling imbalanced data in poverty classification, Stat. J. IAOS 39, 233-239, 2023.
  21. C. Seiffert, T.M. Khoshgoftaar, J. Van Hulse and A. Napolitano, Mining data with rare events: A case study, Proc. 19th IEEE Int. Conf. Tools Artif. Intell., 132-139, 2007.
  22. J.W. Smith, J.E. Everhart, W.C. Dickson, W.C. Knowler and R.S. Johannes, Using the ADAP learning algorithm to forecast the onset of diabetes mellitus, Proc. Symp. Comput. Appl. Med. Care, 261-265, 1988.
  23. S. Smiti and M. Soui, Bankruptcy prediction using deep learning and borderline SMOTE, Inf. Syst. Front. 22, 1067-1083, 2020.
  24. H. Wang, J. Lyu, L. Chen and B. Zhang, Novel ACGAN-based fault detection framework for unbalanced data, Proc. 7th Int. Conf. Adv. Algorithms Control Eng., 705-708, 2024.
  25. J. Wang, M. Xu, H. Wang and J. Zhang, Classification of imbalanced data by using the SMOTE algorithm and locally linear embedding, in: Proc. 8th Int. Conf. Signal Process. 3, IEEE, 2006.
  26. L. Xu, M. Skoularidou, A. Cuesta-Infante and K. Veeramachaneni, Modeling tabular data using conditional GAN, Adv. Neural Inf. Process. Syst. 32, 2019.
  27. G. Zhao, P. Liu, K. Sun, Y. Yang, T. Lan and H. Yang, Research on data imbalance in intrusion detection using CGAN, PLOS ONE 18 (10), e0291750, 2023.
  28. Machine Learning Group - ULB, Credit Card Fraud Detection, Kaggle, 2013. [Online]. Available: https://www.kaggle.com/datasets/mlg-ulb/creditcardfraud. [Accessed: 27-Apr-2026].