Re: Selecting membership functions

On 2007-09-28, Kirk Zurell <kirk@xxxxxxxxxxxxx> wrote:

Thanks for adding this. Working almost exclusively in control,

Kirk,

That was my assumption from both your answer and having looked at your web
site in the past. :-)

... we're aware of the use of smooth curves but have never used them
ourselves (not yet, anyway).

In our approximate reasoning models, they are the predominant shapes.
However, there are plenty of uses for trapezoids and square-shouldered sets
(at the low and high ends of the domain). For example, if the variable is
wetland size, fuzzy term sets for "small," "medium," and "large" can best be
represented by these membership function shapes as most people will agree
that there are ranges of values for each term where the truth value is 1.0.

OP might need these more than the triangles/trapezoids: Inferjack didn't
mention his intended application.

Which is why I mentioned them, with explanations for potential use.

Carpe weekend,

Rich
.