Type-2 fuzzy diagnosis
Peter R. Innocent, Robert John · 2003
This paper describes a computationally fast and simple algorithm that produces an index of support for the diagnosis of diseases based on type-2 fuzzy sets. The procedure allows the specification of fuzzy symptoms by using fuzzy relations to determine a symptom's strength based on its observed intensity and the certainty of its duration. We formally describe the fuzzy approach and the algorithm, and show how it relates to a disease given a range of symptom strengths.