Optimistic Multi-granulation Rough Set-Based Classification for Neonatal Jaundice Diagnosis

Neonatal jaundice diagnosis has been approached by various machine learning techniques. Pattern recognition algorithms are capable of improving the quality of prediction, early diagnosis of diseases, and disease classification. Pattern recognition algorithm results in Neonatal jaundice diagnosis or description of jaundice treatment by the medical specialist. This research focuses on applying rough set-based data mining techniques for Neonatal jaundice data to discover locally frequent identification of jaundice diseases. This work applies Optimistic Multi-granulation rough set model (OMGRS) for Neonatal jaundice data classification. Multi-granulation rough set provides efficient results than single granulation rough set model and soft rough set-based classifier model. The performance of the proposed Multi-granulation rough set-based classification is compared with other Naive bayes, Back Propagation Neural Networks (BPN), and Kth Nearest Neighbor (KNN) approaches using various classification measures.