Computing a family of reproducing kernels for statistical applications (Q2564496)

The author is interested in the study of reproducing kernel Hilbert spaces. At first, it is recalled that Sobolev spaces \(H^m(\Omega)\) equipped with the norm \[ |u|^2=\int u(t)^2dt+ (1/\tau^{2m})\int u^{(m)}(t)^2dt \] constitute a family of such Hilbert spaces. In the introduction some modifications are given. In Section 2 there is established that in the case of the whole real line the kernels are Fourier transforms of simple functions and can be obtained through easy contour integration. In Section 3 the author considers the case of a bounded open interval. At first, the relationship between the kernel on the interval \((a,b)\) associated with \(\tau\) and the kernel on the interval \((0,1)\) associated with another value of this parameter is established. Then, to determine the kernel on \((0,1)\), a system of differential equations of order \(2m\) with initial conditions is derived, and the solution for \(m=1\) and \(m=2\) is given. The Appendix contains a collection of intermediate steps in these computation.