FUZZY LINEAR REGRESSION: AN APPLICATION TO HEART DISEASE

Abstract

Disorders in heart condition refer to heart disease. Several risk factors are associated with causing the heart disease. Physical inactivity and smoking are leading risk factors among other risk factors. The aim of this study is to investigate the relationship of heart disease with physical activity and smoking. Regression analysis is one of the key areas that can be utilized in finding the relationship of variables. By considering heart disease as the output variable (dependent variable) and correlated other factors as input variables, one can model the relationship through multiple linear regression. Fuzzy regression is an application of fuzzy platform for conventional regression analysis. Fuzzy regression analysis gives a fuzzy relationship between dependent and independent variables which represents vagueness in the data. The input data may be crisp values or fuzzy numbers whereas the conventional ordinary least squares regression can handle only crisp measures. The model output is in the form of fuzzy representative which has lower and upper approximation models to represent the fuzziness of the output. Fuzzy models are especially suitable in modelling and predicting heart disease as the disease