integration with respect to a vector measure and function approximation集成对于一个向量测量和函数近似.pdfVIP

INTEGRATION WITH RESPECT TO A VECTOR MEASURE AND FUNCTION APPROXIMATION L. M. GARCÍA-RAFFI, D. GINESTAR, AND E. A. SÁNCHEZ-PÉREZ Received 13 May 2000 The integration with respect to a vector measure may be applied in order to approximate a function in a Hilbert space by means of a finite orthogonal sequence {fi } attending to two different error criterions. In particular, if ∈ R is a Lebesgue measurable set, f ∈ L (), and {A } is a finite family of disjoint subsets of , we can obtain a measure 2 i µ0 and an approximation f0 satisfying the following conditions: (1) f0 is the projection of the function f in the subspace generated by {f } in the Hilbert space f ∈ L (,µ ). i 2 0 (2) The integral distance between f and f on the sets {A } is small. 0 i 1. Introduction Let ⊂ R be a bounded Lebesgue measurable set. We consider the Hilbert space of functions L (). A finite set of orthonormal functions B = {f }n in this space, 2 i i=1 generates a finite-dimensional subspace of L (), . Given a function f ∈ L (), 2 2 we can obtain an approximation of f as the projection of f onto , that is, f can be approximated as n f ≈ P (f ) = α f , (1.1) i i i=1 where the coefficients αi are given by α