6Model Complexity Selection

6.1 Introduction

In black box identification, one of the main problems is choosing the appropriate complexity of the model. This stimulating issue has been studied by scholars of various disciplines, including statisticians, computer scientists, control engineers, and many others. The essential terms of the problem can be outlined as follows. The basic identification criterion adopted so far is


where images is the number of data, images is the parameter vector, and images the prediction error at time images of the considered model images. images provides an index of fit of images to data. Hence, if images is the ...

Get Model Identification and Data Analysis now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.