Take a discretized fuzzy set (i.e. a vector of membership degrees and a vector of numeric values that correspond to that degrees) and perform a selected type of defuzzification, i.e. conversion of the fuzzy set into a single crisp value.
defuzz(
degrees,
values,
type = c("mom", "fom", "lom", "dee", "cog", "expun", "expw1", "expw2")
)A defuzzified value.
A fuzzy set in the form of a numeric vector of membership
degrees of values provided as the values argument.
A universe for the fuzzy set.
Type of the requested defuzzification method. The possibilities are:
'mom': Mean of Maxima - maximum membership degrees are
found and a mean of values that correspond to that degrees is returned;
'fom': First of Maxima - first value with maximum membership
degree is returned;
'lom': Last of Maxima - last value with maximum membership degree
is returned;
'dee': Defuzzification of Evaluative Expressions - method used
by the pbld() inference mechanism that combines the former three
approaches accordingly to the shape of the degrees vector:
If degrees is non-increasing then 'lom' type is used,
if it is non-decreasing then 'fom' is applied, else 'mom' is selected;
'cog': Center of Gravity - the result is a mean of values weighted by degrees;
'exp1': Experimental 1.
Michal Burda
Function converts input fuzzy set into a crisp value. The definition of
input fuzzy set is provided by the arguments degrees and
values. These arguments should be numeric vectors of the same length,
the former containing membership degrees in the interval \([0, 1]\) and
the latter containing the corresponding crisp values: i.e., values[i]
has a membership degree degrees[i].
fire(), aggregateConsequents(), perceive(), pbld(), fcut(), lcut()
# returns mean of maxima, i.e., mean of 6, 7, 8
defuzz(c(0, 0, 0, 0.1, 0.3, 0.9, 0.9, 0.9, 0.2, 0),
1:10,
type='mom')
Run the code above in your browser using DataLab