The Mamdani expert-system with parametric families of fuzzy constraints in evaluation of cancer patient survival length

Abstract

Strict analytic formulas are the tools usually derived for determining the formal relationships between a sample of independent variables and a variable which they affect. If we cannot formalize the function tying the independent and dependent variables then we will utilize some expert-system control actions. We often adopt their fuzzy variants developed by Mamdani, Sugeno and Takagi. Fuzzy expertsystem algorithms are furnished with softer mechanisms, when comparing them to crisp versions. An efficient action of these softer mechanisms depends on the proper fuzzification of variables. At the stage of fuzzifying the variable levels we will prove some parametric expressions, which rearrange one function to several forms needed by the expert-system algorithm. The general parametric equation of membership functions allows creating arbitrary lists without any intuitive assumptions. The fuzzy expert-system algorithms are particularly adaptable to support medical tasks to solve. These tasks often cope with uncertain premises and conclusions. From the medical point of view it would be desirable to prognosticate the survival length for patients suffering from gastric cancer. We thus formulate the objective of the current chapter as the utilization of the Mamdani fuzzy control actions as a methodology adapted for the purpose of making the survival prognoses. © Springer-Verlag Berlin Heidelberg 2013.