Efficacy of Fuzzy-Stat Modelling in Classification of Gynaecologists and Patients

Abstract

Fuzzy logic-based inference systems depend on the domain experts’ perceptions, which are intrinsically imprecise/vague/fuzzy. The perceptions of more than one expert are needed in the decision-making process. Therefore, there is a need to study the similarity between the experts using a mathematical framework. Classical mathematical models simulating the medical diagnostic process are usually either logical or probabilistic, wherein the concept of partial belief is not considered. Except in a few cases, binary logic is too unrealistic to apply to medical diagnosis. Another important factor in medical science is the patient-symptom relationship, which influences the disease diagnosis. In summary, the following two issues stand out: (i) Do experts agree with one another in arriving at the same diagnostic labels? (ii) Based on the symptom-patient relationship, can patients be classified? The authors have tried to explore the possibility of using fuzzy similarity measures and also Gower’s coefficient in classifying gynaecologists and patients. The comparative evaluation infers that the efficacy of two-valued binary logic-based Gower’s coefficient is low.

Keywords

• medical diagnosis
• similarity measures
• cosine amplitude method
• fuzzy set theory
• gower’s coefficient
• max-min method