Fuzzy MCDM
Rachna Jain, Abhishek Kathuria, Devanshi Mukhopadhyay, Meenu Gupta · 2020
In real-world computing, data is raw, abstract, and imprecise, making it difficult to derive results that are fully accurate and reliable. In the field of medicine, data has different levels of uncertainty and imprecision which creates difficulty for physicians to establish the patient&s;s condition from a cluster of similar symptomatic diseases. Doctors diagnose the patient based on statistical assumptions of previous data that the doctor had amassed over years of experience. Further test results and time are required to fully comprehend the patient&s;s condition. Doctors and medical experts may not be equally skilled or have sufficient experience to deal with different kinds of diseases. Furthermore, doctors may not have adequate resources to perform different tests, leading to poor clinical decisions based on intuition alone. Diagnosis also incurs a significant financial burden on the patient. Unlike binary logic, a fuzzy approach works with logic and decision mechanisms that do not have any definite boundaries such as human logic. A physician always has somewhat fuzzy (precise and uncertain) knowledge in the medical field. Hospitals and medical institutions have a huge amount of medical data on different strata of the population, which, if properly classified and used, would help in disease risk prediction. The purpose of applying fuzzy logic with multi-criteria decision-making (MCDM) to the field of medicine is that instead of relying completely on statistical inferences, we also incorporate logic-based assumptions and decisions of the statistical data available. This would enable optimum use of resources and enable the doctor/medical expert to establish the patient&s;s condition accurately and diagnose properly. Further, with disease risk prediction, one could take adequate steps for disease prevention, thus saving time and money. This chapter primarily uses various fuzzy MCDM methods which is used for disease prediction and determination of drug dosage. Medical history of the patient and their family is analyzed and a probabilistic value is generated for the disease that could occur. This chapter also discusses the future scope of neuro-fuzzy applications in detail with end analysis and precise results.