Introduction to Static Models

Abstract

This chapter introduces the basis function formulation, which most subsequent chapters relate to. It systematizes most approaches in one general setting and discusses some very general ideas that are independent of the specific model architecture chosen. These are the notions of linear vs. nonlinear parameters, of local vs. global behavior, and of multiple inputs and/or multiple outputs. Furthermore, the basis function formulation is extended to conveniently include local model networks, which represent a core topic throughout this book. Finally, some artificial test processes are defined, and the different evaluation criteria are discussed.