Minimal value parameters in RKHS
Nadia Souilem, Hassani Messaoud · 2011
In this paper we develop a new algorithm to find the minimal value of the parameter number in the Reproducing Kernel Hilbert Space (RKHS) model which is the number of input / output measurements contained in a learning set. The proposed algorithm consists of computing a criterion for increasing values of this number and repering the value for which the criterion jumps suddenly. The proposed algorithm is validated on a chemical reactor.