Two curve Chebyshev approximation and its application to signal clustering

TL;DR

This paper extends Chebyshev approximation theory to multiple functions, developing a new algorithm for signal clustering that reuses previous results, with theoretical optimality conditions based on nonsmooth convex analysis.

Contribution

It introduces an extension of Chebyshev approximation to multiple functions and proposes an efficient clustering algorithm that leverages previous computations.

Findings

Developed necessary and sufficient optimality conditions for two-curve Chebyshev approximation.

Proposed a new algorithm for signal clustering with result reuse capability.

Extended classical approximation theory to multi-function cases.

Abstract

In this paper we extend a number of important results of the classical Chebyshev approximation theory to the case of simultaneous approximation of two or more functions. The need for this extension is application driven, since such kind of problems appears in the area of curve (signal) clustering. In this paper we propose a new efficient algorithm for signal clustering and develop a procedure that allows one to reuse the results obtained at the previous iteration without recomputing the cluster centres from scratch. This approach is based on the extension of the classical de la Vallee-Poussin's procedure originally developed for polynomial approximation. In this paper, we also develop necessary and sufficient optimality conditions for two curve Chebyshev approximation, that is our core tool for curve clustering. These results are based on application of nonsmooth convex analysis.