Feasibility of random basis function approximators for modeling and control

TL;DR

This paper critically examines the use of random basis function approximators in modeling and control, highlighting their convergence properties and computational demands, and discusses implications for neural network applications.

Contribution

It provides a detailed analysis of the convergence rates of random basis function approximators and discusses their practical limitations in modeling and control tasks.

Findings

Convergence rate of O(1/n) is guaranteed only with substantial computational resources.

Random basis function approximators require significant computational power for effective convergence.

Implications for neural network applications in modeling and control are discussed.

Abstract

We discuss the role of random basis function approximators in modeling and control. We analyze the published work on random basis function approximators and demonstrate that their favorable error rate of convergence O(1/n) is guaranteed only with very substantial computational resources. We also discuss implications of our analysis for applications of neural networks in modeling and control.