Thyroid carcinoma stands as a widespread endocrine malignancy, making the accurate forecast of disease return es-sential for improving post-treatment strategy and patient prognosis. Nevertheless, severe skewness in clinical class distributions and complex feature interactions make predicting recurrence particularly demanding. To address these hurdles, this investigation outlines a robust machine learning architecture incorporating structured preprocessing alongside Bayesian parameter search. The proposed system features duplicate record filtering, categorical label mapping, stratified 80:20 partitioning, SMOTE-Tomek dynamic balancing, and standardization engineered to eliminate data leakage across evaluation folds. Four distinct supervised models—K-Nearest Neighbors (KNN), Support Vector Machines (SVM), Random Forests (RF), and Extreme Gradient Boosting (XGBoost)—were systematically optimized using five-fold stratified cross-validation. Comprehensive benchmarking via Accu-racy, Precision, Sensitivity, F1-Score, Receiver Operating Characteristic curves, and confusion matrix analysis revealed that XGBoost attained superior predictive capacity. Specifically, XGBoost achieved an overall accuracy of 95.89%, precision of 91.30%, sensitivity of 95.45%, F1-score of 93.33%, and ROC-AUC of 97.95%, markedly outperforming the comparative algo-rithms. These outcomes confirm the efficacy of the proposed pipeline in mitigating severe class skew and improving forecasting reliability, establishing its value for clinical decision support and patient follow-up planning.

Our study presented a machine learning methodology for pre-dicting thyroid cancer recurrence from clinical and patholog-ical patient records. Among the evaluated models, XGBoost achieved superior diagnostic accuracy, reaching 95.89% accu-racy, 91.30% precision, 95.45% recall, 93.33% F1-score, and 97.95% ROC-AUC, outperforming KNN, SVM, and Random Forest.

[1]M. Vanderpump, “The epidemiology of thyroid disease,” British Medical Bulletin, vol. 99, no. 1, pp. 39–51, 2011. [2]J. G. Hollowell, “Serum TSH, T4, and thyroid antibodies in the United States population (1988 to 1994): National Health and Nutrition Examination Survey (NHANES III),” Journal of Clinical Endocrinology & Metabolism, vol. 87, no. 2, pp. 489–499, Feb. 2002. [3]F. Pedregosa et al., “Scikit-learn: Machine learning in Python,” Journal of Machine Learning Research, vol. 12, pp. 2825–2830, Nov. 2011. [4]L. Breiman, “Random forests,” Machine Learning, vol. 45, no. 1, pp. 5–32, 2001. [5]A. Raza, F. Eid, E. C. Montero, I. D. Noya, and I. Ashraf, “En-hanced interpretable thyroid disease diagnosis by leveraging synthetic oversampling and machine learning models,” BMC Medical Informatics and Decision Making, vol. 24, no. 1, p. 364, Nov. 2024. [6]A. Tops¸ir, F. Güler, E. Çetin, M. F. Burak, and M. Ag˘raz, “Thy-roid disease classification using generative adversarial net-works and Kolmogorov–Arnold network for three-class classi-fication,” BMC Medical Informatics and Decision Making, vol. 25, no. 1, p. 284, Jul. 2025. [7]A. Hassan et al., “Improving thyroid disorder diagnosis via in-novative stacking ensemble learning model,” Digital Health, vol. 11, Art. no. 20552076251341430, May 2025. [8]S. Akter and H. A. Mustafa, “Analysis and interpretability of machine learning models to classify thyroid disease,” PLoS ONE, vol. 19, no. 5, Art. no. e0300670, May 2024. [9]D. Kumar, D. Raj, A. Singh, T. Rahman, and G. Sharma, “A comprehensive framework for thyroid disorder diagnosis: Inte-grating advanced feature selection, genetic algorithms and ma-chine learning,” PLoS ONE, vol. 20, no. 6, Art. no. e0325900, 2025. [10]D. Kesavulu and R. Kannadasan, “Improved bio-inspired with machine learning computing approach for thyroid prediction,” Scientific Reports, vol. 15, no. 1, p. 22524, Jul. 2025. [11] G. Mollica et al., “Classification of thyroid diseases us-ing machine learning and Bayesian graph algorithms,” IFAC-PapersOnLine, vol. 55, no. 40, pp. 67–72, 2022. [12]V. Kumar et al., “Automated segmentation of thyroid nodule, gland, and cystic components from ultrasound images using deep learning,” IEEE Access, vol. 8, pp. 63482–63496, 2020. [13]V. D. Leitch, J. D. Bassett, and G. R. Williams, “Role of thy-roid hormones in craniofacial development,” Nature Reviews Endocrinology, vol. 16, no. 3, pp. 147–164, 2020. [14]K. Paczkowska et al., “Laboratory interference in the thyroid function test,” Endokrynologia Polska, vol. 71, no. 6, pp. 551–560, 2020. [15]D. Dua and C. Graff, “UCI Machine Learning Repository,” 2017. [Online]. Available: http://archive.ics.uci.edu/ml. [16]S. Razia, P. S. Kumar, and A. S. Rao, “Machine learning tech-niques for thyroid disease diagnosis: A systematic review,” in Modern Approaches in Machine Learning and Cognitive Sci-ence, vol. 885, pp. 203–212, 2020. [17]K. Polat, S. S¸ ahan, and S. Günes¸, “A novel hybrid method based on artificial immune recognition system (AIRS) with fuzzy weighted pre-processing for thyroid disease diagnosis,” Expert Systems with Applications, vol. 32, no. 4, pp. 1141–1147, 2007. [18]A. Akbas, U. Turhal, S. Babur, and C. Avci, “Performance im-provement with combining multiple approaches to diagnosis of thyroid cancer,” Engineering, vol. 5, no. 10, pp. 264–267, 2013. [19]W. Ahmad et al., “Thyroid diseases forecasting using a hybrid decision support system based on ANFIS, k-NN and informa-tion gain method,” Journal of Applied Environmental and Bio-logical Sciences, vol. 7, no. 10, pp. 78–85, 2017. [20]K. S. Parikh et al., “Diagnosing common skin diseases us-ing soft computing techniques,” International Journal of Bio-Science and Bio-Technology, vol. 7, no. 6, pp. 275–286, 2015. [21]L. Ozyilmaz and T. Yildirim, “Diagnosis of thyroid disease us-ing artificial neural network methods,” in Proc. 9th Int. Conf. Neural Information Processing (ICONIP), vol. 4, 2002, pp. 2033–2036. [22]I. Ioni¸ta˘ and L. Ioni¸ta˘, “Prediction of thyroid disease using data mining techniques,” BRAIN: Broad Research in Artificial Intel-ligence and Neuroscience, vol. 7, no. 3, pp. 115–124, 2016. [23]V. Talasila et al., “The prediction of diseases using rough set theory with recurrent neural network in big data analytics,” International Journal of Intelligent Engineering and Systems, vol. 13, no. 5, pp. 10–18, 2020. [24]A. Keles and A. Keles¸, “ESTDD: Expert system for thyroid diseases diagnosis,” Expert Systems with Applications, vol. 34, no. 1, pp. 242–246, 2008. [25]A. Bastias et al., “Predictive model based on neural networks to assist the diagnosis of malignancy of thyroid nodules,” in Proc. 41st Int. Conf. Computers & Industrial Engineering, 2011, pp. 92–97. [26]A. T. Azar, S. A. El-Said, and A. E. Hassanien, “Fuzzy and hard clustering analysis for thyroid disease,” Computer Methods and Programs in Biomedicine, vol. 111, no. 1, pp. 1–16, 2013. [27]D. Chen et al., “A deep-learning based ultrasound text classifier for predicting benign and malignant thyroid nodules,” in Proc. Int. Conf. Green Informatics (ICGI), 2017, pp. 199–204. [28]V. Chaurasia, S. Pal, and B. Tiwari, “Prediction of benign and malignant breast cancer using data mining techniques,” Journal of Algorithms & Computational Technology, vol. 12, no. 2, pp. 119–126, 2018. [29]L. Aversano et al., “Thyroid disease treatment prediction with machine learning approaches,” Procedia Computer Science, vol. 192, pp. 1031–1040, 2021. [30]H. H. Kumar, “A novel approach of SVM based classifica-tion on thyroid disease stage detection,” in Proc. 3rd Int. Conf. Smart Systems and Inventive Technology (ICSSIT), 2020, pp. 836–841. [31]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. [32]I. Tomek, “Two Modifications of CNN,” IEEE Transactions on Systems, Man, and Cybernetics, vol. 6, no. 11, pp. 769–772, Nov. 1976. [33]G. Lemaître, F. Nogueira, and C. K. Aridas, “Imbalanced-learn: A Python Toolbox to Tackle the Curse of Imbalanced Datasets in Machine Learning,” Journal of Machine Learning Research, vol. 18, no. 17, pp. 1–5, 2017. [34]T. M. Cover and P. E. Hart, “Nearest Neighbor Pattern Classifi-cation,” IEEE Transactions on Information Theory, vol. 13, no. 1, pp. 21–27, Jan. 1967. [35]C. Cortes and V. Vapnik, “Support-Vector Networks,” Machine Learning, vol. 20, no. 3, pp. 273–297, 1995. [36]T. Chen and C. Guestrin, “XGBoost: A Scalable Tree Boosting System,” in Proc. 22nd ACM SIGKDD (KDD), San Francisco, CA, USA, 2016, pp. 785–794. [37]J. Snoek, H. Larochelle, and R. P. Adams, “Practical Bayesian Optimization of Machine Learning Algorithms,” in Advances in Neural Information Processing Systems (NeurIPS), vol. 25, 2012. D. M. W. Powers, “Evaluation: From Precision, Recall and F-Measure to ROC, Informedness, Markedness and Correlation,” Journal of Machine Learning Technologies, vol. 2, no. 1, pp. 37–63, 2011.