Enhancing Cluster-Based SMOTE with kNN-Based Post-Oversampling Cleaning for Robust Health Risk Prediction

Mohamad  Ilham; Akhmad  Solikin; Syaad Patmanthara

doi:10.36456/a9k8xq74

Authors

Mohamad Ilham
Akhmad Solikin Universitas PGRI Adi Buana Surabaya
Syaad Patmanthara

DOI:

https://doi.org/10.36456/a9k8xq74

Keywords:

Imbalanced Data, SMOTE, NR-Clustering_SMOTE, epistemology, post-SMOTE cleaning

Abstract

Class imbalance is a common problem in health datasets and often leads to poor recognition of the minority (disease) class. NR-Clustering-SMOTE is a cluster-based oversampling method that combines noise reduction, K-Means clustering, and SMOTE with a modified distance metric to improve classification performance on imbalanced health data. However, the original method only performs noise reduction before SMOTE, so noisy synthetic samples generated around borderline or highly overlapped regions may still degrade classifier performance and introduce epistemological bias in the learned decision rules. This paper proposes a lightweight extension called NR-CluSMOTE-KNC (NR-Clustering-SMOTE with Post-SMOTE k-NN Cleaning). After the standard NR-Clustering-SMOTE pipeline, a k-nearest neighbour filter is applied solely to synthetic minority samples; synthetic points that are surrounded predominantly by majority neighbours are identified as extreme noise and removed. On the Pima Indians Diabetes dataset using Random Forest, the proposed method improves accuracy from 0.8481 (baseline NR-Clustering-SMOTE) to 0.8589 with NR-CluSMOTE-KNC, accompanied by consistent gains in G-Mean and AUC. These results indicate that a simple post-SMOTE cleaning step can epistemologically refine the representation of minority concepts in the data, producing more reliable and fair predictive models for health decision support.

References

[1] M. G. Weiss and F. Provost, “Learning when training data are costly: The effect of class distribution on tree induction,” Journal of Artificial Intelligence Research, vol. 19, pp. 315–354, 2003.

[2] N. V. Chawla, K. W. Bowyer, L. O. Hall, and W. P. Kegelmeyer, “SMOTE: Synthetic minority over-sampling technique,” Journal of Artificial Intelligence Research, vol. 16, pp. 321–357, 2002

[3] H. He and E. A. Garcia, “Learning from imbalanced data,” IEEE Transactions on Knowledge and Data Engineering, vol. 21, no. 9, pp. 1263–1284, Sep. 2009.

[4] A. F. López, B. Krawczyk, M. Woźniak, and N. V. Chawla, “Bringing theory to practice: A comprehensive comparison of induction algorithms for imbalanced data,” Information Sciences, vol. 408, pp. 1–24, Oct. 2017.

[5] M. H. Nguyen, C. K. Kwoh, S. L. Hoi, and C. S. Ong, “A new clustering-based oversampling method for learning from imbalanced data,” in Proc. 5th Int. Conf. on Advanced Data Mining and Applications (ADMA), 2009, pp. 146–159

[6] H. Hairani, T. Widiyaningtyas, D. D. Prasetya, and A. Aminuddin, “Addressing Imbalance in Health Datasets: A New Method NR-Clustering SMOTE and Distance Metric Modification,” Comput. Mater. Contin., vol. 82, no. 2, pp. 2931–2949, 2025.

[7] H. Han, W.-Y. Wang, and B.-H. Mao, “Borderline-SMOTE: A new over-sampling method in imbalanced data sets learning,” in Proc. Int. Conf. on Intelligent Computing (ICIC), Lecture Notes in Computer Science, vol. 3644, 2005, pp. 878–887.

[8] H. He, Y. Bai, E. A. Garcia, and S. Li, “ADASYN: Adaptive synthetic sampling approach for imbalanced learning,” in Proc. IEEE Int. Joint Conf. on Neural Networks (IJCNN), 2008, pp. 1322–1328.

[9] D. L. Wilson, “Asymptotic properties of nearest neighbor rules using edited data,” IEEE Transactions on Systems, Man, and Cybernetics, vol. SMC-2, no. 3, pp. 408–421, 1972.

[10] G. E. A. P. A. Batista, R. C. Prati, and M. C. Monard, “A study of the behavior of several methods for balancing machine learning training data,” SIGKDD Explorations, vol. 6, no. 1, pp. 20–29, 2004.

[11] H. Hairani, T. Widiyaningtyas, and D. D. Prasetya, “Revue d ’ Intelligence Artificielle Feature Selection and Hybrid Sampling with Machine Learning Methods for Health Data Classification,” vol. 38, no. 4, pp. 1255–1261, 2024.

[12] P. Zhao, X. Liu, Z. Yue, Q. Zhao, X. Liu, and Y. Deng, “Computer Methods and Programs in Biomedicine Update DiGAN Breakthrough : Advancing diabetic data analysis with innovative GAN-based imbalance correction techniques,” Comput. Methods Programs Biomed. Updat., vol. 5, no. December 2023, p. 100152, 2024.

[13] A. Islam, S. Brahim, A. Ur, and H. Bensmail, “KNNOR : An oversampling technique for imbalanced datasets,” Appl. Soft Comput., vol. 115, p. 108288, 2022.

[14] E. Blanco-mallo, L. Morán-fernández, B. Remeseiro, and V. Bolón-canedo, “Do all roads lead to Rome ? Studying distance measures in the context of machine learning,” Pattern Recognit., vol. 141, p. 109646, 2023.

[15] S. Maldonado, J. López, and C. Vairetti, “An alternative SMOTE oversampling strategy for high-dimensional datasets,” Appl. Soft Comput. J., vol. 76, pp. 380–389, 2019.

[16] N. Islam, S. A. Turza, S. I. Fahim, and R. M. Rahman, “International Journal of Cognitive Computing in Engineering Advanced Parkinson ’ s Disease Detection : A comprehensive artificial intelligence approach utilizing clinical assessment and neuroimaging samples,” Int. J. Cogn. Comput. Eng., vol. 5, no. May, pp. 199–220, 2024.

[17] A. Arafa, N. El-fishawy, M. Badawy, and M. Radad, “RN-SMOTE : Reduced Noise SMOTE based on DBSCAN for enhancing imbalanced data classification,” J. King Saud Univ. - Comput. Inf. Sci., vol. 34, no. 8, pp. 5059–5074, 2022.

[18] A. V. D. Sano, F. M. Bhatti, E. Miranda, M. Aryuni, A. Y. Zakiyyah, and C. Bernando, “The Impact of Augmentation and SMOTE Implementation on the Classification Models Performance: A Case Study on Student Academic Performance Dataset,” Procedia Comput. Sci., vol. 245, pp. 282–289, 2024.

[19] S. Matharaarachchi, M. Domaratzki, and S. Muthukumarana, “Machine Learning with Applications Enhancing SMOTE for imbalanced data with abnormal minority instances,” Mach. Learn. with Appl., vol. 18, no. October, p. 100597, 2024.

[20] S. Szeghalmy and A. Fazekas, “Knowledge-Based Systems A comparative study on noise filtering of imbalanced data sets,” Knowledge-Based Syst., vol. 301, no. June, p. 112236, 2024.

[21] Y. Liu, W. Zhang, G. Qin, and J. Zhao, “^ ĐŝĞŶĐĞ ŝƌĞĐƚ ScienceDirect 9th International Conference on Information Technology and Quantitative Management A comparative study on the effect of data imbalance on software A comparative study on defect the effect of data imbalance on software prediction defect prediction,” Procedia Comput. Sci., vol. 214, pp. 1603–1616, 2022.

[22] S. Shi, W. Luo, and G. Pau, “An attention-based balanced variational autoencoder method for credit card fraud detection,” Appl. Soft Comput., vol. 177, p. 113190, 2025.

[23] H. Zhou, J. Tong, Y. Liu, K. Zheng, and C. Cao, “Journal of King Saud University - Computer and Information Sciences An oversampling FCM-KSMOTE algorithm for imbalanced data classification,” J. King Saud Univ. - Comput. Inf. Sci., vol. 36, no. 10, p. 102248, 2024.

[24] N. Ashidah, A. Abdullah, and N. Mat, “Association features of smote and rose for drug addiction relapse risk,” J. King Saud Univ. - Comput. Inf. Sci., vol. 34, no. 9, pp. 7710–7719, 2022.

[25] Z. Farou, Y. Wang, and T. Horváth, “Intelligent Systems with Applications Cluster-based oversampling with area extraction from representative points for class imbalance learning,” Intell. Syst. with Appl., vol. 22, no. August 2023, p. 200357, 2024.

[26] Mukhtar, M. Nushron Ali, et al. "Analisis desain stator generator tipe magnet permanen fluks aksial menggunakan metode finite element analysis (FEA)." ELEMEN: JURNAL TEKNIK MESIN 8.2 (2021): 149-156.

[27] Atmiasri, Atmiaasri, and Sagita Rochman. "Pendeteksi Logam Untuk Industri Makanan Berbasis PLC." WAKTU: Jurnal Teknik UNIPA 9.1 (2011): 78-81.