The HDM: Support for Darwin, Fuzzy-Logic or the Neural Model?

B. Bagot (Germany)



This paper explores some reflections, properties, a new formula and further generalizations based on the “harmonic decision matrix” (HDM), a recent fuzzy operator and neural model. First considering the decision function from the point of view of evolutionary algorithms, fuzzy-logic and neural model, we give then an insight into its use in a complex real world environment. Through chosen examples and comparisons, we underline the role of the engineer at key decision levels in the use of soft computing programs. We point out the advantages of the HDM as parameterized, user defined selection mechanism and help to the decision that clearly continues a series of more and more complex aggregation functions. The HDM turns its back on Pareto-front techniques in multiobjective optimizations, replaces the true/false opposition by good/bad results in the Fuzzy-logic and enhances the role of single neurons in neural nets. Therefore, as it deals with a new paradigm [18] in these disciplines, this paper must be strongly philosophical.

Important Links:

Go Back