An Optimized Soft Computing Framework with Rough Set Min-Max Classifier for Medical Data Diagnosis

Abstract

Correct diagnosis of medical problems is one of the most critical elements of health care, requiring the use of advanced methods of studying medical data sets that are growing in complexity. The Rough Set Min-Max Classifier (RMMC) is a new approach which utilizes the rough set theory with the goal of feature selection and classification. This article introduces the RMMC (Regional Marine Planning). In order to find key characteristics, decrease duplicate attributes, and improve the effectiveness of classification, the RMMC makes use of a unique neighbouring connection that is centred on Euclidean distance. The framework provides high prediction result accuracy with the use of the RMMC on real-world datasets from the healthcare related fields such as the Neurological Seizures Classification and the Breast Cancer datasets. This helps the algorithm to work well with complex data sets.  Comparisons of the results with classic classifiers like the K-Nearest Neighbours (KNN) algorithm, the Support Vector Machines (SVM) algorithm and the Neighbourhood Rough Set Classifier (NRSC) have demonstrated that RMMC has succeeded significantly over them. The robustness of the framework has been validated using k-fold cross validation further adding to its capability of becoming a reliable diagnostic tool in medical information analytics. The RMMC establishes a new standard for healthcare diagnosis thanks to its proved capacity to enhance making choices and minimize the intricacy of computing processes. Given its precision, recall and f1-score values of 0.998, 0.95 and 0.98, respectively, it proves to be effective in attaining 100% accuracy in its forecast.