Нечіткі логічні висновки і висновки в експертних системах медичної діагностики

• Дата: 2021
• Автор: Zack, Yuriy
• Формат: Стаття
• Мова: Російська
• Опубліковано: The National Technical University of Ukraine "Igor Sikorsky Kyiv Polytechnic Institute" 2021
• Теми: кластерний аналіз; багатовимірні функції належності; центроїди fuzzy-множин об’єктів і кластерів; центри тяжкості і середини перетинів нечітких множин; критерії оптимальності та алгоритми кластеризації
• Онлайн доступ: http://journal.iasa.kpi.ua/article/view/244471
• Назва журналу: System research and information technologies

The main problems in making a correct diagnosis are: subjectivity and insufficient qualifications of the doctor, difficulties in correctly assessing the patient’s complaints, signs and symptoms of the disease observed in the patient, as well as individual manifestations of the symptoms of the disease. In publications on the use of expert systems for medical diagnostics using fuzzy logic, the main attention was paid to the medical features of the problem. In this work, for the first time, general methodological aspects of building such systems, creating databases, representing by fuzzy sets of real numbers, digital scales, linguistic and Boolean data of symptom values are formulated. The types of membership functions that are advisable to use to represent the symptoms of diseases are proposed. In fuzzy-logical conclusions, not only the values of the characteristic functions of the logical terms of individual symptoms, but also complex arithmetic functions of their values are used.