Diagnosis of spinal column pathologies using ensemble with rejection option
DOI:
https://doi.org/10.59681/2175-4411.v16.iEspecial.2024.1216Keywords:
Cervical Column, Confidence, Reject OptionAbstract
Objective: To propose a new approach for decision-making with rejection option in classifier committees. Method: The developed method encompasses classification techniques using ensembles with the Rejection Option approach, employing the Gini Index (GI) as a confidence metric. We established thresholds based on the distribution of purity percentages obtained from each class, allowing the model to abstain from predicting samples that are difficult to classify in medical diagnostics related to cervical column diseases. Results: The proposed model outperformed comparisons, achieving 97.55% accuracy and rejecting 61.69% of samples in the most conservative scenario. The Accuracy and Rejection curve highlighted its superiority. Conclusion: Defining GI value ranges offers flexibility in adjusting the committee's rigidity, further revealing potential for optimizing classifier committees in various applications, providing greater reliability in pattern recognition.
References
Reshi AA, Ashraf I, Rustam F, Shahzad HF, Mehmood A, Choi GS. Diagnosis of vertebral column pathologies using concatenated resampling with machine learning algorithms. PeerJ Comput Sci. 22 de julho de 2021;7:e547.
Rocha Neto AR, Sousa R, de A. Barreto G, Cardoso JS. Diagnostic of Pathology on the Vertebral Column with Embedded Reject Option. Em: Vitrià J, Sanches JM, Hernández M, organizadores. Pattern Recognition and Image Analysis. Berlin, Heidelberg: Springer; 2011. p. 588–95.
Nanglia S, Ahmad M, Ali Khan F, Jhanjhi NZ. An enhanced Predictive heterogeneous ensemble model for breast cancer prediction. Biomed Signal Process Control. 1o de fevereiro de 2022;72:103279.
Zhang XY, Xie G, Li XC, Mei T, Liu CL. A Survey on Learning to Reject. Proc IEEE. 2023.
Mienye ID, Sun Y. A Survey of Ensemble Learning: Concepts, Algorithms, Applications, and Prospects. IEEE Access. 2022;10:99129–49.
Chow C. On optimum recognition error and reject tradeoff. IEEE Trans Inf Theory. 1970.
Bartlett PL, Wegkamp MH. Classification with a Reject Option using a Hinge Loss. J Mach Learn Res. 1o de junho de 2008;9:1823–40.
Fukunaga K. Introduction to statistical pattern recognition. 2. ed. San Diego [u.a.]: Acad. Press; 1990.
Dubuisson B, Masson M. A statistical decision rule with incomplete knowledge about classes. Pattern Recognit. 1o de janeiro de 1993;26(1):155–65.
Hellman ME. The Nearest Neighbor Classification Rule with a Reject Option. IEEE Trans Syst Sci Cybern. julho de 1970;6(3):179–85.
Cordella LP, Foggia P, Sansone C, Tortorella F, Vento M. Classification reliability and its use in multi-classifier systems. Em: Del Bimbo A, organizador. Image Analysis and Processing. Berlin, Heidelberg: Springer; 1997. p. 46–53.
Villon S, Mouillot D, Chaumont M, Subsol G, Claverie T, Villéger S. A new method to control error rates in automated species identification with deep learning algorithms. Sci Rep. 2020.
Guilherme Barreto AN. Vertebral Column [Internet]. UCI Machine Learning Repository; 2005. Disponível em: https://archive.ics.uci.edu/dataset/212
Ferreira AJ, Figueiredo MAT. Boosting Algorithms: A Review of Methods, Theory, and Applications. 2012.
Chen T, Guestrin C. XGBoost: A Scalable Tree Boosting System. Em: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York, NY, USA: Association for Computing Machinery; 2016. p. 785–94. Disponível em: https://dl.acm.org/doi/10.1145/2939672.2939785
Homenda W, Luckner M, Pedrycz W. Classification with rejection based on various SVM techniques. Em: 2014 International Joint Conference on Neural Networks (IJCNN). Beijing, China: IEEE; 2014. p. 3480–7. Disponível em: https://ieeexplore.ieee.org/document/6889655
Tan PN, Steinbach M, Karpatne A, Kumar V. Introduction to Data Mining. Pearson; 2019.
Gamelas Sousa R, Rocha Neto AR, Cardoso JS, Barreto GA. Robust classification with reject option using the self-organizing map. Neural Comput Appl. 1o de outubro de 2015;26(7):1603–19.
Anand V, KiranBala B, Srividhya S, C. K, Younus M, Rahman MH. Gaussian Naïve Bayes Algorithm: A Reliable Technique Involved in the Assortment of the Segregation in Cancer. Mob Inf Syst. 17 de junho de 2022;2022:1–7.
Ballabio D, Todeschini R, Consonni V. Chapter 5 - Recent Advances in High-Level Fusion Methods to Classify Multiple Analytical Chemical Data. Em: Cocchi M, organizador. Data Fusion Methodology and Applications. Elsevier; 2019. p. 129–55. Disponível em: https://www.sciencedirect.com/science/article/pii/B9780444639844000053
Dogan A, Birant D. A Weighted Majority Voting Ensemble Approach for Classification. Em: 2019 4th International Conference on Computer Science and Engineering (UBMK). 2019. p. 1–6.
Yuan Y, Wu L, Zhang X. Gini-Impurity Index Analysis. IEEE Trans Inf Forensics Secur. 2021.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2024 Reginaldo Pereira Fernandes Ribeiro, Ajalmar Rego da Rocha Neto, Thiago Alves Rocha
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
Submission of a paper to Journal of Health Informatics is understood to imply that it is not being considered for publication elsewhere and that the author(s) permission to publish his/her (their) article(s) in this Journal implies the exclusive authorization of the publishers to deal with all issues concerning the copyright therein. Upon the submission of an article, authors will be asked to sign a Copyright Notice. Acceptance of the agreement will ensure the widest possible dissemination of information. An e-mail will be sent to the corresponding author confirming receipt of the manuscript and acceptance of the agreement.
