Operator means of probability measures (Q2308295): Difference between revisions

The study of operator means began with the work of \textit{W. N. Anderson jun.} and \textit{R. J. Duffin} [J. Math. Anal. Appl. 26, 576--594 (1969; Zbl 0177.04904)] who first studied the arithmetic and harmonic means and proved the arithmetic-harmonic mean inequality. The geometric mean was introduced by \textit{W. Pusz} and \textit{S. L. Woronowicz} [Rep. Math. Phys. 8, 159--170 (1975; Zbl 0327.46032)]. \textit{T. Ando} [Linear Algebra Appl. 26, 203--241 (1979; Zbl 0495.15018)] proved the arithmetic-geometric mean inequality. The axiomatic theory for connections and means for pairs of positive operators was developed by \textit{K. Nishio} and \textit{T. Ando} [J. Math. Anal. Appl. 53, 539--549 (1976; Zbl 0332.94010)] and \textit{F. Kubo} and \textit{T. Ando} [Math. Ann. 246, 205--224 (1980; Zbl 0412.47013)]. A~binary operation \(\sigma \) on positive invertible operators on a Hilbert space is called operator connection if it satisfies certain axioms, namely, monotonicity, transformer inequality, and upper continuity. If \(\sigma \) further satisfies the normalization condition \(I\sigma I\), then it is called operator mean. Kubo and Ando [loc.\,cit.]\ showed the existence of an affine order isomorphism between the class of connections and the class of positive operator monotone functions on \(R_+\). This isomorphism \(\sigma \longleftrightarrow f\) is characterized by the relation \(A\sigma B=A^{1/2}f(A^{-1/2}BA^{-1/2})A^{1/2}\). In this paper, the authors introduce the notion of operator means of probability measures on the set of positive invertible operators on a Hilbert space, in accordance with Kubo and Ando's definition of operator means. They show that every operator mean is contractive for the \(\infty\)-Wasserstein distance. Deformation of such operator means is considered and it is shown that the deformation of any operator mean becomes again an operator mean. Using this deformation procedure, they prove a number of properties and inequalities for operator means of probability measures.