This paper proposes a multi-objective approach to designing a new class of hybrid intelligent systems in the field of medical diagnostics. The knowledge base construction of the intelligent system is based on a hybrid model of an evidential classifier, founded on combining fuzzy logistic regression, which possesses interpretability properties, with Dempster-Shafer evidence combination technology. This technology allows for combining the classifier’s results obtained from multiple information sources and integrating specialists’ knowledge within a single hybrid system. To improve the efficiency and performance of the evidential classifier, a mechanism for calibrating the feature scale of the log-regression model is proposed. An example of calibrating a log-model describing the results of a 6-min test for diagnosing chronic heart failure is provided. The decision-making mechanism in the hybrid system utilizes the Dempster-Shafer independent evidence combination technology. To ensure the correct application of Shafer’s technology for information fusion, a formal definition of evidence independence is introduced, and a criterion to satisfy this condition is proposed. The article discusses the generalized architecture of a hybrid intelligent system for solving medical diagnostic tasks, as well as the most important elements of the multi-objective design technology for such systems.
DOI: 10.1007/978-3-032-13612-1_33
Download the proceedings from Springer Nature: https://link.springer.com/content/pdf/10.1007/978-3-032-13612-1.pdf
Kobrinsky, B. A., Kovalev, S. M., Chekanova, V. S. (2026). Hybrid Approach to Designing Medical Intelligent Systems Based on Combining Fuzzy Models and Heterogeneous Information Fusion Methods // Proceedings of the Ninth International Scientific Conference “Intelligent Information Technologies for Industry” (IITI’25), Volume 2, pp. 372–385.