Classification of imbalanced lung cancer stage data using wang’s boosted support vector machine

Karisma, Ria Dhea Layla Nur ORCID: https://orcid.org/0000-0002-5941-9565, Ameliya, Dhea P., Pagalay, Usman, Widayani, Heni ORCID: https://orcid.org/0000-0002-6966-6754, Saadati, Hawzah and Jamhuri, Mohammad ORCID: https://orcid.org/0000-0002-4506-7669 (2026) Classification of imbalanced lung cancer stage data using wang’s boosted support vector machine. Presented at 2026 International Conference on Artificial Intelligence Robotics, Signal and Image Processing (AIRoSIP), 06 August 2026, Yogyakarta.

Text 29428.pdf - Published Version Restricted to Repository staff only Download (948kB)

Abstract

Imbalanced data frequently occurs in medical data classification and can negatively impact model performance, particularly for minority classes. The missing values in the dataset may further affect classification outcomes. This study used Wang’s Boosted Support Vector Machine (WBSVM) through a combination of Support Vector Machine (SVM), boosting techniques, and asymmetric misclassification cost that assigns greater weights to the misclassification of minority classes. Missing values were handled using mode imputation before the modeling process was carried out. The data used in this study consist of medical records of lung cancer patients at RSUD Karsa Husada Batu from 2020-2025 with variables including age, sex, smoking status, comorbidities, family history, and cancer stage. The research stages include data preprocessing and stratified sampling using 5-fold cross- validation with a composition of 64% training data, 16% validation data, and 20% testing data. The model was developed using a linear kernel with parameters λ=1, C=1, and the number of iterations M=40. The result shows that the model has been a sensitivity value of 0.5714, specificity of 0.6017, and G-mean of 0.5864. It indicates that the model achieved a moderate balance in recognizing majority and minority classes. However, the model performance may reflect the limitations of the data set and the linear kernel.

Item Type:	Conference (Paper)
Keywords:	asymmetric cost; g-mean; imbalanced data; imputation; minority class; wang’s boosted support vector machine
Subjects:	01 MATHEMATICAL SCIENCES > 0104 Statistics > 010401 Applied Statistics 01 MATHEMATICAL SCIENCES > 0104 Statistics > 010402 Biostatistics
Divisions:	Faculty of Mathematics and Sciences > Department of Mathematics
Depositing User:	Miss Ria Dhea Layla Nur Karisma
Date Deposited:	03 Nov 2026 08:45

Downloads

Origin of downloads

Actions (login required)

View Item