Fuzzy Expert System to Diagnose Diabetes Using S Weights for S Fuzzy Assessment Methodology

Introduction

Fuzzy Logic is an efficient tool to deal with vagueness and uncertainties. Heart disease is one of the diseases which is spread around the world. It suddenly kills the human community. In recent trends fuzzy logic and expert system plays an efficient role to solve this problem. Fuzzy expert system is a user friendly approach for reasoning and controlling with the presence of uncertainty for variety of problem domains. The Fuzzy Expert System defines imprecise knowledge and offers linguistic concept with excellent approximation. Many researchers are working in the area of fuzzy expert system to make a correct decision or to diagnosis the problem. Many fuzzy expert systems has been developed in all field ie., in medicine, agriculture, industries and computer field. Most emerging and flouring technique is fuzzy expert system used in the field of medicine. In the field of medicine the use of fuzzy expert system is highly increased to diagnosis disease. Medical practitioners exhibit variation in decision making to deal with uncertainties and ambiguity in knowledge and information. In the field of medicine there are many high risk diseases which kill the human community. They are heart disease, diabetes, high blood pressure and cancer. Diabetes and heart disease is the high risk disease that arises day by day in India and around the world. So Fuzzy Expert System is very much essential to develop an early detection and assessment tool for patients who suffer from diseases. Medical diagnosis needs careful examination of patients, to make a decision regarding the patients whether they suffer from some disease or not.

In modern world the use of computers and its technology has been increased in the field of medicine for diagnosis, for the treatment of illnesses to the patients is highly increased. To handle this type of situation use of intelligent systems such as fuzzy logic, fuzzy expert system, knowledge-based system, statistical technique, fuzzy model, Bayesian statistics, artificial neural network and genetic algorithm have been developed. The expert system has been implemented on the web application due to the rapid changes of the Information Communication Technology (ICT). The very important benefits of the web application are that expert system, helps to deliver the real expertise to the people who are at the remote location or the people who needs occasionally expertise. Diagnostic expert systems are the prime example for web based application. Elevator Company is a good example for online expert system. An elevator company contracts with the customer to provide maintenance, with the design of online expert system, which helps to determine whether the problem identified needs a technician to make a service call or the customer is capable to solve the problem themselves without the technicians. Online expert system saves the customers time and money.

Diabetes Mellitus is a heterogeneous group of diseases, which lead to high blood glucose levels due to defects in either insulin secretion or insulin action. A hybrid method is designed which is used for fuzzy weighted pre-processing and Artificial Immune Recognition System (AIRS). Artificial Immune Recognition System shows an effective performance on several problems such as machine learning benchmark problems and medical classification problems like breast cancer, diabetes, liver disorders classification (Polat & Sahan, 2005). Graphical User Interface (GUI) is designed to enter the patient’s record and predict whether the patient is having heart disease. The prediction is performed with data mining techniques by using historical data or data repository. In Weighted Associative Classifier (WAC), different weights are assigned to different attributes according to their predicting capability. Classifiers are performing well than traditional classifiers approaches such as decision tree and rule induction. WAC is used to generate rules. The system is implemented in java platform and dataset used is UCI machine learning repository (Soni, Ansari, & Sharma, 2011).