Open Access

ARTICLE

Addressing Class Overlap in Sonic Hedgehog Medulloblastoma Molecular Subtypes Classification Using Under-Sampling and SVD-Enhanced Multinomial Regression

Isra Mohammed¹, Mohamed Elhafiz M. Musa², Murtada K. Elbashir^3,*, Ayman Mohamed Mostafa³, Amin Ibrahim Adam⁴, Mahmood A. Mahmood³, Areeg S. Faggad⁵

1 Department of Statistics, Faculty of Mathematical and Computer Sciences, University of Gezira, Wad Madani, 21113, Sudan
2 Department of Computer Science, College of Computer and Information Sciences, Jouf University, Sakaka, 72388, Saudi Arabia
3 Department of Information Systems, College of Computer and Information Sciences, Jouf University, Sakaka, 72388, Saudi Arabia
4 Department of Statistics, Faculty of Economic and Social Studies, Omdurman Islamic University, Khartoum, 11111, Sudan
5 Department of Molecular Biology, National Cancer Institute, University of Gezira, Wad Madani, 21113, Sudan

* Corresponding Author: Murtada K. Elbashir. Email:

Computers, Materials & Continua 2025, 84(2), 3749-3763. https://doi.org/10.32604/cmc.2025.063880

Received 27 January 2025; Accepted 29 April 2025; Issue published 03 July 2025

Abstract

Sonic Hedgehog Medulloblastoma (SHH-MB) is one of the four primary molecular subgroups of Medulloblastoma. It is estimated to be responsible for nearly one-third of all MB cases. Using transcriptomic and DNA methylation profiling techniques, new developments in this field determined four molecular subtypes for SHH-MB. SHH-MB subtypes show distinct DNA methylation patterns that allow their discrimination from overlapping subtypes and predict clinical outcomes. Class overlapping occurs when two or more classes share common features, making it difficult to distinguish them as separate. Using the DNA methylation dataset, a novel classification technique is presented to address the issue of overlapping SHH-MB subtypes. Penalized multinomial regression (PMR), Tomek links (TL), and singular value decomposition (SVD) were all smoothly integrated into a single framework. SVD and group lasso improve computational efficiency, address the problem of high-dimensional datasets, and clarify class distinctions by removing redundant or irrelevant features that might lead to class overlap. As a method to eliminate the issues of decision boundary overlap and class imbalance in the classification task, TL enhances dataset balance and increases the clarity of decision boundaries through the elimination of overlapping samples. Using fivefold cross-validation, our proposed method (TL-SVDPMR) achieved a remarkable overall accuracy of almost 95% in the classification of SHH-MB molecular subtypes. The results demonstrate the strong performance of the proposed classification model among the various SHH-MB subtypes given a high average of the area under the curve (AUC) values. Additionally, the statistical significance test indicates that TL-SVDPMR is more accurate than both SVM and random forest algorithms in classifying the overlapping SHH-MB subtypes, highlighting its importance for precision medicine applications. Our findings emphasized the success of combining SVD, TL, and PMR techniques to improve the classification performance for biomedical applications with many features and overlapping subtypes.

---

Keywords

---