A Comparative Analysis of Feature Selection Approaches on Base and Ensemble Machine Learning Models in Breast Cancer Diagnosis

Beyza Kalcı; Betül Uzbaş; Ahmet Babalık

doi:10.54047/bibted.2025412

EN TR

A Comparative Analysis of Feature Selection Approaches on Base and Ensemble Machine Learning Models in Breast Cancer Diagnosis

Abstract

In this study, combinations of machine learning and ensemble models using different feature selection techniques were examined on the Breast Cancer Wisconsin dataset, and the differences in performance of these models across various feature selection techniques were compared. Logistic Regression (LR), Support Vector Machines (SVM), Multilayer Perceptron (MLP), k-Nearest Neighbor (k-NN), Decision Trees (DT), Random Forest (RF), Naive Bayes (NB), XGBoost, LightGBM and ensemble learning methods were used in the study. Two different feature selection methods (Filter and Wrapper) were applied to optimize model performance. To prevent data leakage, feature selection operations were performed only on the training data. To validate the models, 5-fold cross-validation was performed with 10 different seed values. The performance of the predictions was evaluated using Accuracy, Precision, Recall, Specificity, F1 Score, Matthews Correlation Coefficient (MCC) and ROC-AUC. The models with the highest accuracy obtained from various combinations of feature selection and machine learning were analyzed using the nonparametric Friedman test, and the difference between the two most successful models was evaluated using the Nadeau & Bengio-corrected t-test. Finally, the decision-making process of the most successful model was explained using SHAP (Shapley Additive exPlanations) analysis for transparency and interpretability.

Keywords

Breast Cancer Diagnosis, Ensemle Learning, Machine Learning, Feature Selection, Classification

Meme Kanseri Teşhisinde Özellik Seçimi Yaklaşımlarının Temel ve Topluluk Makine Öğrenmesi Modelleri Üzerindeki Karşılaştırmalı Analizi

Abstract

Bu çalışmada, Breast Cancer Wisconsin veri seti üzerinde farklı özellik seçimi teknikleri ile makine öğrenmesi ve topluluk modellerinin kombinasyonları ele alınmış ve bu modellerin farklı özellik seçimi tekniklerine karşı gösterdiği performans farkları karşılaştırılmıştır. Çalışmada Lojistik Regresyon (LR), Destek Vektör Makineleri (SVM), Çok Katmanlı Algılayıcı (MLP), En Yakın Komşu (k-NN), Karar Ağaçları (DT), Rastgele Orman (RF), Naive Bayes (NB), XGBoost, LightGBM ve topluluk öğrenme yöntemleri kullanılmıştır. Model performansını optimize etmek için iki farklı özellik seçimi yöntemi uygulanmıştır. (Filtre ve Sarmal). Veri sızıntısı önlemek amacıyla özellik seçimi işlemleri sadece eğitim verisi üzerinde yapılmıştır. Modellerin doğrulanması için 10 farklı seed değeriyle 5 katlı çapraz doğrulama yapılmıştır. Tahminlerin performansları Doğruluk, Kesinlik, Geri Çağırma, Özgüllük, F1 Puanı, Matthews Korelasyon Katsayısı (MCC) ve ROC-AUC kullanılarak değerlendirilmiştir. Farklı öznitelik seçimi ve makine öğrenmesi kombinasyonlarından elde edilen en yüksek doğruluğa sahip modeller non-parametrik Friedman testi ile analiz edilmiş, öne çıkan en başarılı iki model arasındaki fark ise Nadeau & Bengio düzeltmeli t-testi ile değerlendirilmiştir. Son olarak da şeffaflık ve yorumlanabilirlik için en başarılı modelin karar verme süreci SHAP (Shapley Additive exPlanations) analizi kullanılarak açıklanmıştır.

Keywords

Meme Kanseri Teşhisi, Özellik Seçimi, Sınıflandırma, Topluluk Öğrenmesi, Makine Öğrenmesi

References

Arravalli, T., Chadaga, K., Muralikrishna, H. et al. Detection of breast cancer using machine learning and explainable artificial intelligence. Sci Rep 15, 26931 (2025). doi:10.1038/s41598-025-12644-w
Arrieta, A. B., Díaz-Rodríguez, N., Del Ser, J., Bennetot, A., Tabik, S., Barbado, A., ... Herrera, F. (2020). Explainable artificial intelligence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible AI. Information Fusion, 58, 82–115. doi:10.1016/j.inffus.2019.12.012
Bozdağ Ak, D., & Selvi, İ. H. (2025). Prediction of Dye Removal Using Machine Learning Techniques. Sakarya University Journal of Computer and Information Sciences, 8(3), 496-509. doi:10.35377/saucis...1697738
Breiman, L. Random Forests. Machine Learning 45, 5–32 (2001). doi:10.1023/A:1010933404324
Breiman, L., Friedman, J. H., Olshen, R. A., & Stone, C. J. (1984). Classification and regression trees. Belmont, CA: Wadsworth International Group.
Chandrashekar, G., & Sahin, F. (2014). A survey on feature selection methods. Computers & Electrical Engineering, 40(1), 16–28. doi:10.1016/j.compeleceng.2013.11.024
Chen, T., & Guestrin, C. (2016). XGBoost: A scalable tree boosting system. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (pp. 785–794). San Francisco, CA. doi:10.1145/2939672.2939785
Cortes, C., Vapnik, V. Support-vector networks. Mach Learn 20, 273–297 (1995). doi:10.1007/BF00994018
Cover, T. M., & Hart, P. E. (1967). Nearest neighbor pattern classification. IEEE Transactions on Information Theory, 13(1), 21–27. doi:10.1109/TIT.1967.1053964
Dhahri H, Al Maghayreh E, Mahmood A, Elkilani W, Faisal Nagi M. Automated Breast Cancer Diagnosis Based on Machine Learning Algorithms. J Healthc Eng. 2019 Nov 3;2019:4253641. doi: 10.1155/2019/4253641

Eftekharian, M., Hosseiny, M., Motallebi, N. & Norouzkhani Esterabadi, S. (2025). Accurate and Interpretable Breast Cancer Diagnosis Using Logistic Regression: An Evaluation on the Wisconsin Diagnostic Dataset. InfoScience Trends, 2(9), 52–60. doi:10.61882/ist.202502.09.04
Garba, A.T., Hamza, H.S. Interpretable Machine Learning Approach for Breast Cancer Classification. Hum-Cent Intell Syst 5, 308–322 (2025). doi:10.1007/s44230-025-00111-8
Guyon, I., & Elisseeff, A. (2003). An introduction to variable and feature selection. Journal of Machine Learning Research, 3, 1157–1182.
Hand, D. J., & Yu, K. (2001). Idiot's Bayes—Not so stupid after all? International Statistical Review, 69(3), 385–398. doi:10.1111/j.1751-5823.2001.tb00465.x
Hashim, M. S., & Yassin, A. A. (2023). Using Pearson Correlation and Mutual Information (PC-MI) to select features for accurate breast cancer diagnosis based on a soft voting classifier. Iraqi Journal for Electrical and Electronic Engineering, 19(2), 43–53. doi: 10.37917/ijeee.19.2.6
Hosmer, D. W., Lemeshow, S., & Sturdivant, R. X. (2013). Applied logistic regression (3rd ed.). Hoboken, NJ: John Wiley & Sons. doi:10.1002/9781118548387
Houssami N, Skaane P. Overview of the evidence on digital breast tomosynthesis in breast cancer detection. Breast. 2013 Apr;22(2):101-108. doi: 10.1016/j.breast.2013.01.017
Ke, G., Meng, Q., Finley, T., Wang, T., Chen, W., Ma, W., Ye, Q., & Liu, T.-Y. (2017). LightGBM: A highly efficient gradient boosting decision tree. In Advances in Neural Information Processing Systems (Vol. 30, pp. 3146–3154).
Aqeel Ahmed Khan, & Muhammad Abu Bakr. (2024). Enhancing Breast Cancer Diagnosis with Integrated Dimensionality Reduction and Machine Learning Techniques. Journal of Computing & Biomedical Informatics, 7(02). Retrieved from https://jcbi.org/index.php/Main/article/view/573
T. Khater et al., "An Explainable Artificial Intelligence Model for the Classification of Breast Cancer," in IEEE Access, vol. 13, pp. 5618-5633, 2025, doi: 10.1109/ACCESS.2023.3308446
Kohavi, R., & John, G. H. (1997). Wrappers for feature subset selection. Artificial Intelligence, 97(1–2), 273–324. doi:10.1016/S0004-3702(97)00043-X
Lundberg, S. M., & Lee, S.-I. (2017). A unified approach to interpreting model predictions. In I. Guyon, U. V. Luxburg, S. Bengio, H. Wallach, R. Fergus, S. Vishwanathan, & R. Garnett (Eds.), Advances in neural information processing systems (Vol. 30, pp. 4765–4774).
Mitchell, T. M. (1997). Machine learning. New York, NY: McGraw-Hill.
Murty, P. S. R. C., Anuradha, C., Naidu, P. A., Mandru, D., Ashok, M., Rajeswaran, N., & Saravanan, V. (2024). Integrative hybrid deep learning for enhanced breast cancer diagnosis: Leveraging the Wisconsin Breast Cancer Database and the CBIS-DDSM dataset. Scientific Reports, 14, 26287. doi:10.1038/s41598-024-74305-8
Nadeau, C., & Bengio, Y. (2003). Inference for the generalization error. Machine Learning, 52(3), 239–281. doi:10.1023/A:1024068626366
Rumelhart, D. E., Hinton, G. E., & Williams, R. J. (1986). Learning representations by back-propagating errors. Nature, 323(6088), 533–536. doi:10.1038/323533a0
Sandhu, J. K., Sharma, C., Kaur, A., Pandey, S. K., Sinha, A., & Shreyas, J. (2025). Development of a clinical decision support system for breast cancer detection using ensemble deep learning. Scientific Reports, 15, 26098. doi:10.1038/s41598-025-06784-2
Sung, H., Ferlay, J., Siegel, R. L., Laversanne, M., Soerjomataram, I., Jemal, A., & Bray, F. (2021). Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA: A Cancer Journal for Clinicians, 71(3), 209–249. doi:10.3322/caac.21660
Wolberg, W. H., Street, W. N., & Mangasarian, O. L. (1994). Machine learning techniques to diagnose breast cancer from fine-needle aspirates. Cancer Letters, 77(2–3), 163–171. doi:10.1016/0304-3835(94)90099-X
Zhu, J., Zhao, Z., Yin, B., Wu, C., Yin, C., Chen, R., & Ding, Y. (2025). An integrated approach of feature selection and machine learning for early detection of breast cancer. Scientific Reports, 15, 13015. doi:10.1038/s41598-025-97685-x

Details

Primary Language

English

Subjects

Machine Learning (Other)

Journal Section

Research Article

Authors

Beyza Kalcı ^*
0009-0007-4959-286X
Türkiye

Betül Uzbaş
0000-0002-0255-5988
Türkiye

Ahmet Babalık
0000-0003-0146-4221
Türkiye

Publication Date

October 27, 2026

Submission Date

August 25, 2026

Acceptance Date

October 12, 2026

Published in Issue

Year 2026 Volume: 6 Number: 2

DOI

https://doi.org/10.54047/bibted.2025412

IZ

https://izlik.org/JA43GT69CP

APA

Kalcı, B., Uzbaş, B., & Babalık, A. (2026). A Comparative Analysis of Feature Selection Approaches on Base and Ensemble Machine Learning Models in Breast Cancer Diagnosis. Bilgisayar Bilimleri Ve Teknolojileri Dergisi, 6(2). https://doi.org/10.54047/bibted.2025412

AMA

1.Kalcı B, Uzbaş B, Babalık A. A Comparative Analysis of Feature Selection Approaches on Base and Ensemble Machine Learning Models in Breast Cancer Diagnosis. BIBTED. 2026;6(2). doi:10.54047/bibted.2025412

Chicago

Kalcı, Beyza, Betül Uzbaş, and Ahmet Babalık. 2026. “A Comparative Analysis of Feature Selection Approaches on Base and Ensemble Machine Learning Models in Breast Cancer Diagnosis”. Bilgisayar Bilimleri Ve Teknolojileri Dergisi 6 (2). https://doi.org/10.54047/bibted.2025412.

EndNote

Kalcı B, Uzbaş B, Babalık A (October 1, 2026) A Comparative Analysis of Feature Selection Approaches on Base and Ensemble Machine Learning Models in Breast Cancer Diagnosis. Bilgisayar Bilimleri ve Teknolojileri Dergisi 6 2

IEEE

[1]B. Kalcı, B. Uzbaş, and A. Babalık, “A Comparative Analysis of Feature Selection Approaches on Base and Ensemble Machine Learning Models in Breast Cancer Diagnosis”, BIBTED, vol. 6, no. 2, Oct. 2026, doi: 10.54047/bibted.2025412.

ISNAD

Kalcı, Beyza - Uzbaş, Betül - Babalık, Ahmet. “A Comparative Analysis of Feature Selection Approaches on Base and Ensemble Machine Learning Models in Breast Cancer Diagnosis”. Bilgisayar Bilimleri ve Teknolojileri Dergisi 6/2 (October 1, 2026). https://doi.org/10.54047/bibted.2025412.

JAMA

1.Kalcı B, Uzbaş B, Babalık A. A Comparative Analysis of Feature Selection Approaches on Base and Ensemble Machine Learning Models in Breast Cancer Diagnosis. BIBTED. 2026;6. doi:10.54047/bibted.2025412.

MLA

Kalcı, Beyza, et al. “A Comparative Analysis of Feature Selection Approaches on Base and Ensemble Machine Learning Models in Breast Cancer Diagnosis”. Bilgisayar Bilimleri Ve Teknolojileri Dergisi, vol. 6, no. 2, Oct. 2026, doi:10.54047/bibted.2025412.

Vancouver

1.Beyza Kalcı, Betül Uzbaş, Ahmet Babalık. A Comparative Analysis of Feature Selection Approaches on Base and Ensemble Machine Learning Models in Breast Cancer Diagnosis. BIBTED. 2026 Oct. 1;6(2). doi:10.54047/bibted.2025412

Aim & Scope

The main aim of the Journal of Computer Science and Technologies is to contribute to academic knowledge by publishing original and high-quality scientific studies in the fields of computer science, software engineering, artificial intelligence, data science, cybersecurity, information technologies, and related disciplines.

The journal seeks to strengthen scientific communication among researchers, academicians, industry professionals, and students, while encouraging the dissemination of current technological developments and supporting interdisciplinary studies.

All submissions are evaluated through a Double-Blind Peer Review process in accordance with the principles of scientific impartiality and publication ethics.

The scientific, ethical, and legal responsibility of the published articles belongs entirely to the authors.

The following types of manuscripts are accepted in our journal:

Research Article: Scientific studies presenting original research findings.
Review Article: Comprehensive studies reviewing the literature on a specific subject.
Industrial Article: Studies focusing on industrial applications, innovative products, and technological developments.
Thesis-Based Study: Articles derived from graduate theses and adapted into scientific manuscript format.
Short Paper: Studies presenting preliminary findings, innovative ideas, or concise research outputs.
Book Review/Comment: Evaluative writings on recent scientific books relevant to the field.

The Journal of Computer Science and Technologies covers theoretical, applied, and interdisciplinary studies in the fields of computer science, software, and information technologies.

The scope of the journal includes the following areas:

Artificial Intelligence and Intelligent Systems
Machine Learning
Data Science and Data Mining
Algorithms and Theory of Computation
Computer Networks
Cybersecurity
Cryptography
Computer Architecture
Operating Systems
Parallel and Distributed Systems
Computer Graphics
Computer Vision
Image and Signal Processing
Natural Language Processing
Scientific Computing
Modeling and Simulation
Mobile and Cloud Computing Systems
Bioinformatics
Medical Informatics
Information Retrieval
Human–Computer Interaction
Software Engineering

The journal welcomes innovative and scientifically significant studies addressing current developments in computer science and technologies, without being limited to the fields listed above.

Author Guidelines

General Principles for Authors

Manuscripts submitted to the Journal of Computer Science and Technologies must meet at least one of the following criteria:

Research studies presenting original scientific findings,
Studies addressing practical applications through scientific methods and approaches,
Review studies comprehensively evaluating current developments in a particular field.

Submitted manuscripts must not have been previously published or simultaneously submitted to another journal or publication for evaluation.

It is assumed that all authors have agreed upon the order of authorship in the manuscript. The scientific, ethical, and legal responsibility for published studies belongs entirely to the authors.

Published studies, in whole or in part, may not be used without proper citation. Manuscripts may not be published elsewhere or presented as conference papers without the permission of the Editorial Board.

Submission of Manuscripts

Manuscripts must be submitted electronically through the DergiPark submission system. Submission and upload procedures should be completed via the following address:

https://dergipark.org.tr/tr/journal/3293/submission/start

Manuscript Preparation Guidelines

Manuscripts must be prepared in accordance with the journal’s formatting and writing guidelines. Authors are encouraged to use the sample manuscript template during the preparation process.

📝 Sample Manuscript File (docx)

Documents Required for Manuscript Submission

Mandatory Documents:

Plagiarism Report
As of 2017, all manuscripts submitted to the journal are required to be uploaded together with a plagiarism report (iThenticate, Turnitin, etc.).
Manuscripts without a plagiarism report may be rejected by the Editorial Board without further evaluation. In addition, submitted manuscripts may be re-examined using iThenticate or similar software when deemed necessary by the Editorial Board.
Copyright Transfer Form
The Copyright Transfer Form must be completed, signed, and uploaded to the system in PDF format.
The signed form must be scanned and uploaded to the system. Manuscripts submitted without this document cannot be published.
📝 Copyright Transfer Form (doc)

Optional Documents:

Reviewer Suggestion Form
Authors may upload a Reviewer Suggestion Form during the submission process in order to contribute to the evaluation procedure. The Editorial Board reserves the full right to decide whether or not to consider the proposed reviewers during the review process.
📝 Reviewer Suggestion Form (doc)

Contact

Prof. Dr. Erdinç AVAROĞLU
Editor-in-Chief

E-mail: eavaroglu@mersin.edu.tr

Ethical Principles and Publication Policy

PUBLICATION ETHICS

The publication processes applied in the Journal of Computer Science and Technologies constitute the basis for the development and distribution of information in an impartial and respectful manner. The processes implemented in this direction are directly reflected in the quality of the work of the authors and the institutions that support the authors. Peer-reviewed studies are studies that embody and support the scientific method. At this point, it is important that all stakeholders of the process (authors, readers and researchers, publishers, referees and editors) comply with the standards of ethical principles. All stakeholders are expected to bear the following ethical responsibilities within the scope of the publication ethics of the Journal of Computer Science and Technologies.

The ethical missions and responsibilities below have been prepared by taking into account the guidelines and policies published by the Committee on Publication Ethics (COPE) as open access.
The editors of the Journal of Computer Science and Technologies have the following ethical missions and responsibilities on the basis of the "COPE Code of Conduct and Best Practice Guidelines for Journal Editors" and "COPE Best Practice Guidelines for Journal Editors" published by the Committee on Publication Ethics (COPE) must be.

Plagiarism İnspection;

The Journal of Computer Science and Technologies will check plagiarism for all submitted articles before they are previously published. If plagiarism content is detected by editors, the author will be notified for rewriting. Plagiarism will be prevented by scanning the entire article by iThenticate®. In any case, if more than 20% of the original article submitted is plagiarism, then the article will be rejected and the author will also be notified. We strongly recommend that you check your content before submitting it to the Journal of Computer Science and Technologies for publication. While plagiarism is not entirely reliable, some free plagiarism programs are recommended.

Ethical Responsibilities of Authors;

1- The works submitted by the author (s) are expected to be original. If the author (s) benefit from other studies or use other studies, they must reference and / or cite completely and accurately.
2- People who do not contribute to the content intellectually in the creation of the study should not be specified as a writer.
3- The situations and relationships that may constitute a conflict of interest, if any, should be explained.
4- Raw data regarding the articles can be requested from the author (s) within the framework of the evaluation processes; In such a case, the author (s) should be ready to present the expected data and information to the editorial board.
5- The author (s) must have the document showing that they have the right to use the data used and the necessary permissions for research / analysis.
6- If the author (s) notices an error or error related to his published, early-view or evaluation phase, there is an obligation to inform the editor of the journal or the publisher, to cooperate with the editor in correction or withdrawal.
7- Authors cannot keep their works in the process of application of more than one journal at the same time. Each application can be started following the completion of the previous application. The study published in another journal cannot be sent to the Journal of Computer Science and Technologies.

Ethical Responsibilities of Referees;

1- It should only agree to evaluate studies of its area of expertise.
2- It should evaluate in neutrality and confidentiality.
3- If he thinks that he is facing a conflict of interest during the evaluation process, he should refuse to review the study and inform the journal editor.
4- In accordance with the principle of confidentiality, it should destroy the studies they have examined after the evaluation process. They can only use the final versions of the studies they have studied only after their publication.
5- It should make the assessment objectively only in relation to the content of the study. Nationality should not allow gender, religious beliefs, political beliefs, and commercial concerns to influence evaluation.
6- He should make the assessment in a constructive and kind language. It should not make humiliating personal comments that include hostility, slander and insults.
7- They should carry out the studies they accepted to evaluate in a timely manner and with the above ethical responsibilities.

Ethical Responsibilities of Editors;

1- All publications are evaluated based on the intellectual content of the author, regardless of gender, race, ethnicity, religion, citizenship and political values.
2- This information about the articles is kept confidential.
3- The conflict of interest observed for the article should be disclosed.
4- While the editorial board evaluates the article, the journal is based on editorial board policies, legally prohibited articles against plagiarism, insults and copyright infringement and takes responsibility for decisions to publishing for submitted articles.

Price Policy

Indexes

Journal Boards

Editor

Erdinç AVAROĞLU

Information and Computing Sciences, Cryptography, Engineering, Electrical Engineering

Editorial Board or Advisory Board

AA

Ali ARI

Artificial Intelligence (Other)

ÖT

Ömer TÜRK

Image Processing, Deep Learning, Machine Vision , Information Security and Cryptology

ES

Esra SARAÇ

Machine Learning Algorithms, Machine Learning (Other), Information Retrieval and Web Search, Natural Language Processing, Evolutionary Computation, Artificial Intelligence (Other)

SY

Selman YAKUT

Algorithms and Calculation Theory, Information Security and Cryptology, Cryptography, Data and Information Privacy, Computer Software

TT

Taner TUNCER

Image Processing, Deep Learning

Kemal ADEM

Image Processing, Artificial Intelligence (Other)

Faruk SERIN

Information and Computing Sciences, Image Processing, Machine Learning Algorithms, Natural Language Processing, Artificial Intelligence (Other), Engineering

İsmail KOYUNCU

Embedded Systems

JB

Jale BEKTAŞ

Image Processing, Computational Imaging, Pattern Recognition

Technical Editor

OA

Oben ADAK

Image Processing, Deep Learning, Machine Learning Algorithms, Natural Language Processing

Language Editor

JB

Jale BEKTAŞ

Image Processing, Computational Imaging, Pattern Recognition

A Comparative Analysis of Feature Selection Approaches on Base and Ensemble Machine Learning Models in Breast Cancer Diagnosis

A Comparative Analysis of Feature Selection Approaches on Base and Ensemble Machine Learning Models in Breast Cancer Diagnosis

Abstract

Keywords

Meme Kanseri Teşhisinde Özellik Seçimi Yaklaşımlarının Temel ve Topluluk Makine Öğrenmesi Modelleri Üzerindeki Karşılaştırmalı Analizi

Abstract

Keywords

References

Details

Primary Language

Subjects

Journal Section

Authors

Publication Date

Submission Date

Acceptance Date

Published in Issue

DOI

IZ

Cite

Aim & Scope

Author Guidelines

General Principles for Authors

Submission of Manuscripts

Manuscript Preparation Guidelines

Documents Required for Manuscript Submission

Contact

Ethical Principles and Publication Policy

Price Policy

Indexes

Other Indexes

Journal Boards

Editor

Editorial Board or Advisory Board

Technical Editor

Language Editor