Introductory chapter: ANFIS for modeling multivariable nonlinear functions
Constantin Voloşencu · IntechOpen eBooks · 2024
The approximation errors vary a lot on the surface of the nonlinear function, being on certain portions small, acceptable, but on other portions quite large, having to be reduced through other attempts.It is likely that, by doing other trainings, a structure can be obtained that ensures, maybe, smaller approximation errors, but there is no such sure certainty.