A Novel Soft Computing Method Based on Interval Type-2 Fuzzy Logic for Classification of Celiac Disease

Abstract

This study proposes a new method for automatic classification of medical data on celiac disease (CD) using an interval type-2 fuzzy logic system (IT2FLS). Due to the high uncertainty in the medical data, IT2FLSs are able to consider many linguistic uncertainties in the type-2 framework, and considering the uncertainty in the membership functions, they can raise the accuracy of the fuzzy system. To improve the performance of IT2FLS, we have used fuzzy C-means (FCM) clustering algorithm to determine the membership functions centers in the fuzzy rules. For the purpose of comparison, other classification models based on fuzzy sets, such as type-1 fuzzy logic system (T1FLS) and IT2FLS without using FCM are also proposed. Experiments are performed on a dataset of Poursina Hakim Research Institute with the real samples of patients with different grades of celiac. Accuracy of 83.33, 87.88 and 90.65, respectively, was achieved when determining the grades of A, B1 and B2 by IT2FLS-FCM. This demonstrates the superiority of this model over the other fuzzy models. Considering the uncertainty in type-2 fuzzy sets and as well as FCM clustering algorithm, improved system performance in the classification of CD.