Quasi-invariant measures on commutative Lie groups of infinite product type (Q910084): Difference between revisions
From MaRDI portal
Added link to MaRDI item. |
Set profile property. |
||
Property / MaRDI profile type | |||
Property / MaRDI profile type: MaRDI publication profile / rank | |||
Normal rank |
Revision as of 02:35, 5 March 2024
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Quasi-invariant measures on commutative Lie groups of infinite product type |
scientific article |
Statements
Quasi-invariant measures on commutative Lie groups of infinite product type (English)
0 references
1991
0 references
Let \(G^ N\) be the infinite product of a finitely generated commutative Lie group G. For a probability \(\mu\) on \(G^ N\), put \(Y_{\mu}=\{a\in G^ N\); \(\tau_ a\mu \sim \mu \}\) \((=\) the quasi-invariant translations of \(\mu)\), where \(\tau_ a\mu =\mu (\cdot -a)\). Using the convergence of infinite products of independent random variables and the Fourier transforms on G, we give several estimations for \(Y_{\mu}\) of power probabilities and exchangeable probabilities on \(G^ N\) in terms of some \(\ell_ p\)-type spaces, i.e. \(\ell_ p(G,H)=\{a=(a_ k)\in G^ N\); \(\sum_{k\in N}| a_ k|^ p_ H<\infty \}\) where H is a subgroup of G and \(| \cdot |_ H\) is a quotient metric on G/H. It turns out that the p-th moments of the adjoint probability of \(\mu\) (which is defined via Fourier transform) are crucial to estimate \(Y_{\mu}\). As an application we treat the quasi-invariance of power and exchangeable probabilities on \(R^ N\) and \(C^ N\) with respect to diagonal matrices. The p-summable matrices work as our yardstick for measurement of the deviation between quasi-invariant matrices and measure-preserving ones.
0 references
finitely generated commutative Lie group
0 references
convergence of infinite products of independent random variables
0 references
exchangeable probabilities
0 references
quasi-invariant
0 references
measure-preserving
0 references