On convergence of conditional probability measures

From MaRDI portal





In an information-theoretical framework, the author discusses some relationships between the convergence of probability measures conditioned on sample mean events and certain minimum I-divergences (resp. Kullback- Leibler information numbers) under constraints. Convergence is obtained in the total variation metric as well as in the \(\tau\)-topology. The results are based upon Sanov-type large deviation theorems due to \textit{P. Groeneboom}, \textit{J. Oosterhoff} and \textit{F. H. Ruymgaart} [Ann. Probab. 7, 553-586 (1979; Zbl 0425.60021)]. For a comprehensive study on the convergence in information of conditional probability measures, we also refer to \textit{I. Csiszár} [ibid. 12, 768-793 (1984; Zbl 0544.60011)].











This page was built for publication: On convergence of conditional probability measures

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1080267)