Abstract:
Early diagnosis of Diabetes is important as it reduces the chances of related complications to
arise. Swarm Intelligence is being widely used for medical diagnostic purposes. Many classifiers
are being optimized by Swarm Intelligence techniques to reduce the cumbersome procedure of
defining complex rules and frameworks. Cuckoo Search is a recently developed algorithm which
uses the concept of Swarm Intelligence. Cuckoo Search mimics the brooding behavior of some
Cuckoo species. Cuckoo Search is enhanced by Levy Flights which follows Levy distribution.
Fuzzy Reasoning Model represents some expert knowledge via simple linguistic rules. This
makes the system more understandable. Employing Fuzzy Reasoning Model to represent a
system for diabetes diagnosis will help an expert who has to consider a large number of factors
before making a decision. In this way human error will be minimized.
Objective of this thesis is to employ Fuzzy Reasoning Model for Diabetes Diagnosis. Cuckoo
Search has been used to optimize fuzzy model for better results. Pima Indians Diabetes Data set
has been used to evaluate the accuracy of proposed classifier.
