A fuzzy classification system based on Ant Colony Optimization for diabetes disease diagnosis

Classification systems have been widely utilized in medical domain to explore patient’s data and extract a predictive model. This model helps physicians to improve their prognosis, diagnosis or treatment planning procedures. The aim of this paper is to use an Ant Colony-based classification system to extract a set of fuzzy rules for diagnosis of diabetes disease, named FCS-ANTMINER. We will review some recent methods and describe a new and efficient approach that leads us to considerable results for diabetes disease classification problem. FCS-ANTMINER has new characteristics that make it different from the existing methods that have utilized the Ant Colony Optimization (ACO) for classification tasks. The obtained classification accuracy is 84.24% which reveals that FCS-ANTMINER outperforms several famous and recent methods in classification accuracy for diabetes disease diagnosis.

► This paper is the first comprehensive proposal for utilizing ACO for Fuzzy Rule Discovery from diabetes disease data.
► FCS-ANTMINER has new features that make it different from existing ACO-Based classification systems.
► The accuracy of FCS-ANTMINER is 84.24% for diabetes disease data which is superior to several famous and recent classifiers.
► FCS-ANTMINER produces a few short fuzzy rules which reveal the interpretability of the resulted classification system.