Skip to content

DCT Lexicon · Neuro-fuzzy

Membership function (graded declaration)

μ

The possibility a graded declaration (fuzzy set) gives each event, from 0 to 1; its cuts are nested crisp sets.

Vol. II, Glossary of Key Terms, p. 399, verbatim without its citations

In depth

Formal statement

A graded declaration (a fuzzy set) gives each event a possibility μ(ξ)∈[0,1]\mu(\xi) \in [0, 1], its membership; its α\alpha-cut Uα={ξ:μ(ξ)≥α}U_\alpha = \{\xi : \mu(\xi) \ge \alpha\} holds the events of possibility at least α\alpha. The possibility of a set of events EE is Π(E)=sup⁡Eμ\Pi(E) = \sup_E \mu, and the necessity of its complement is 1−Π(E)1 - \Pi(E).

Vol. II, Definition 10.1, verbatim

Executive example

A graded declaration is how offices actually speak. “Deviations are usually within $50 million of plan and never beyond $200 million” is a membership that is 1 up to $50 million and 0 beyond $200 million; each cut is a crisp bound, from the usual to the unthinkable.

Vol. II, Section 10.1, verbatim

Common misreading

A common mistake is to read a possibility as a probability.

Vol. II, LOS 10.3, verbatim

How AXIOM uses it

In the Readiness (ANFIS) tab each answer is a triangular fuzzy number on the 1–5 scale, and the scale's five labels are Gaussian terms. Each strip shades levels 1 to 5 by how firmly that dimension's answers support them. The shades are possibilities, not the shares of a population.

Drafted for the author's approval, from AXIOM's ANFIS engine (D-093)

In the book: Vol. II, Ch. 10: Neuro-Fuzzy Robust Enterprise Optimization

Related: ANFIS (adaptive neuro-fuzzy inference) · Robust optimization (max-min plan)

V2.Def.10.1 · V2.Thm.10.2 · V2.LOS.10.3