A sparse method for the canonical representation and approximation of PWL functions
Pedro Marcelo Julian, M. Pique, Osvaldo Enrique Agamennoni · 2002
The purpose of this paper is to show that an orthonormal formulation of the high level canonical representation previously proposed [see IEEE Trans. Circuits Syst. I, vol. 46, pp. 463-480, April 1999], together with the use of sparse matrix and vector techniques, permits one to efficiently formulate and handle large scale numerical problems.