A Comparative Analysis of Machine Learning Techniques for Identification and Categorization of Dermatological Disorders

The integration of intelligent computing and machine learning (ML) has transformed various fields, leading to significant advancements in automation, data analysis, and decision-making processes. The diagnosis of diseases and access to healthcare services have long been in high demand, yet they face significant challenges. Machine learning offers a wide range of effective solutions for this purpose. While skin disorders may appear harmless, they can become serious if not treated properly. Among the most challenging issues in medical image processing is identifying skin conditions from clinical photos. Furthermore, identifying skin conditions by hand by medical professionals is subjective and time-consuming. Therefore, automatic skin disease prediction is necessary for both patients and dermatologists, which speeds up treatment planning. In this paper a comparitive study has been done on various classification algorithms and performed on various healthcare datasets and the most effective classification algorithm for the dermatological dataset has been identified. Classification algorithms have been compared on the basis of their classification accuracy. The study suggested that Logistic Model Trees (LMT) and Navie Bayes Algorithms can be used for the efficient classification of dermatology disorders.