How Lebesgue decomposition works Part 1 (Machine Learning 2024) | Written by Monodeep Mukherjee | April 2024

Machine Learning


Monodeep Mukherjee
Photo by Riho Kroll on Unsplash
  1. Non-commutative Lebesgue decomposition of non-commutative measures (arXiv)

Author: Fuad Naderi

Summary: A positive noncommutative (NC) measure is a positive linear function on a free disk operator system produced by a d-tuple of noncommutative isometry. We construct a natural Lebesgue decomposition for positive NC measures relative to other such measures by introducing a hybrid form, its Cauchy transform, and the technique of NC reproduction kernel Hilbert space (RKHS). Our study is an extension of the Jury-Martin decomposition that originally decomposed the positive NC measure against the standard NC Lebesgue measure. In fact, it provides a more generalized definition of absolute continuity and singularity. This comes down to the definition when the split measure is a standard NC Lebesgue measure. This generalized definition makes it possible to extend the Jury-Martin theory for arbitrary split NC measures, recovering the decomposition when the split NC measure is a Lebesgue measure. Our study implies a Lebesgue decomposition for the representation of the Kunz-Toeplitz C* algebra. Moreover, our RKHS method gives a new proof of the classical Lebesgue decomposition when applied to the classical one-dimensional setting, i.e. d = 1.

2. Representation of mapping of semi-bounded form and its Lebesgue-type decomposition (arXiv)

Author: Seppo Hassi, Henk de Snou

Summary: For a semibounded sesquilinear form t in a Hilbert space H, there exists a representation mapping Q from H to another Hilbert space K.[φ,ψ]−c(φ,ψ)=(Qφ,Qψ), φ,ψ∈domt, c∈R is the lower bound of t. The map representation provides a simplification tool for studying general semi-bounded forms. By representing the map, the closedness, closure, and singularity of t are immediately translated into the corresponding properties of the operator Q, and vice versa. Also, the properties of the sum decomposition t=t1+t2 between a non-negative form t and two other non-negative forms t1 and t2 in H can be analyzed by the associated non-negative reduction K∈B(K). This helps, for example, to establish an explicit operator theory characterization of whether the addends t1 and t2 are mutually singular. Such a sum decomposition is used to study the properties of the so-called Lebesgue-type decomposition of the semibounded form t. Here t1 is closable and t2 is singular. In particular, this includes the Lebesgue decomposition of semi-bounded forms by B. Simon. Furthermore, for a semi-bounded form t with a representation mapping Q, the corresponding semi-bounded self-adjoint relation Q∗Q∗∗+c is uniquely defined by the limit version of the classical representation theorem for the form t studied by W. is shown to be determined. Arendt and T. ter Erst by field. Representing the map completely handles the convergence of monotonic sequences in semibounded form.



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *