Evaluation machine-learning approaches for classification of Cryotherapy and Immunotherapy datasets

Machine-learning (ML) methods have great importance when applied interdisciplinary. Besides many areas, ML methods save cost and time in medical applications. In this study, we experimented several ML methods with different approaches on classification of Cryotherapy and Immunotherapy datasets, which are applied on wart treatment. The effects of dimension reduction techniques and handling of unbalanced sample classes are the main discussion points of our study. When several ML models are analyzed, Random Forest (RF) achieved 95% accuracy, %88 sensitivity, and %98 specificity. Other ML methods also performed successful results close to the RF. Although some promising results were obtained, we also discussed the drawbacks of these approaches while evaluating wart treatment strategies.