| MATLAB Function Reference | Search  Help Desk |
| identify |
identifies a fuzzy model from data
Syntax
m=identify(m,in,out,options)
Description
m=identify(m,in,out,options)
identify model m on the basis of the data
inside in and out using method specified in options.
The accepted fields for options are:
method identification method(see below).
n_rules number of rules.
min_n_rules min. nb. of rules (incremental methods only)
max_n_rules max. nb. of rules (incremental methods only)
fuzzyness fuzzyness of the model (usually greater than 1)
tolerance minimal improvement for recursive techniques before the algorithm stops.
seed seed for random generator
lin_fit type of linear fitting: 0: normal, 1: local
rule_type rule type code (see add_rules for more info)
cluslms GK clustering followed by a LMS fit
fmclust Robert Babuska's identification method
incrsie Incremental identification method (Siemens)
tsgklmxv incremental method developped in IRIDIA
cluslev GK clust. init. and Lev.-Marq. optimisation
randlm random init. and Lev.-Marq. optimisation