The homogeneous chaos from the standpoint of vector measures (Q4353716)

From MaRDI portal

!

This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:

scientific article; zbMATH DE number 1059578
Language Label Description Also known as
default for all languages
No label defined
    English
    The homogeneous chaos from the standpoint of vector measures
    scientific article; zbMATH DE number 1059578

      Statements

      The homogeneous chaos from the standpoint of vector measures (English)
      0 references
      0 references
      26 May 1999
      0 references
      Wiener's equality
      0 references
      homogeneous chaos
      0 references
      Einstein's Brownian motion stochastic process
      0 references
      non-Gaussian random variables
      0 references
      covariance structure
      0 references
      Lebesgue-Pettis integrability condition
      0 references
      Fubini type theorem
      0 references
      tensor products
      0 references
      Brownian motion random measures
      0 references
      control measure problem
      0 references
      Lebesgue type decompositions
      0 references
      integration with respect to Markovian kernels
      0 references
      The author develops a theory of chaotic measures and chaotic integration starting from the basic contribution about ``The homogeneous chaos'' by \textit{N. Wiener} [Am. J. Math. 60, 897-936 (1938; Zbl 0019.35406)]. The main ingredients of N. Wiener's contribution concerns the notion of the \(p\)-fold product \(\xi_p\) over \(\mathbb{R}^p\) of his original measure \(\xi\) associated with his idealized Einstein's Brownian motion stochastic process, where the range of \(\xi_p\) consists of square-integrable, non-Gaussian random variables with an involved covariance structure. The complicated covariance structure reflects the difficulty of a proof for versions of Wiener's equality concerning the expectation of \(\xi(A_1)\cdot\cdots\cdot \xi(A_{2n})\) for Borel subsets \(A_1,\dots, A_{2n}\) of the real line with compact closure, which are derived here by means of involved combinatorial arguments. In particular, the Lebesgue-Pettis integrability condition \(f\in{\mathcal P}_{1,\xi_p}\), where \({\mathcal P}_{1,\xi_p}\) stands for the closure of simple functions relative to the \(\xi_p\)-norm, yields automatically the existence of liftings appearing in the work of \textit{Y. Z. Hu} and \textit{P. A. Meyer} about ``Chaos de Wiener et intégrale de Feynman'' [Lect. Notes Math. 1321, 51-71 (1988; Zbl 0644.60081)]. Moreover, a Fubini type theorem for tensor products like \(f\times g\in{\mathcal P}_{1,\xi_{p+ q}}\), \(f\in{\mathcal P}_{1,\xi_q}\) \((p,q\in \{1,2,\dots\})\) has to be shown for the validity of Wiener's equation. In this connection, it is quite illuminating to observe that all mathematical difficulties would have not appeared if N. Wiener would have introduced \(\xi_p\) by tensor products instead of sticking to just one \(L_2\)-space for all \(p\in\{1,2,\dots\}\). The author starts his theory on chaotic measures by proving the existence of Brownian motion random measures according to \textit{J. P. Kahane} [``Some random series of functions'', Heath Math. Monographs (1968; Zbl 0192.53801)] based on linear isometries. Moreover, the countable additivity of \(\xi_p\) is proved, which follows also from Theorem 4.5 appearing in Mem. Am. Math. Soc. 265, 82 p. (1982; Zbl 0489.60064) of \textit{D. D. Engel} about ``The multiple stochastic integral'', where, however, no Wiener type equation is proved. Moreover, it is shown that the measure \(\xi_p\) is for \(p\geq 2\) not absolutely continuous with respect to the \(p\)-dimensional Lebesgue measure and product structures, the control measure problem as well as Lebesgue type decompositions are investigated in connection with the measure \(\xi_p\). Furthermore, an inversion formula for \(\xi_p\) in the spirit of \textit{Gian-Carlo Rota's} work ``On the foundations of combinatorial theory. I: Theory of Möbius functions'' [Z. Wahrscheinlichkeitstheorie verw. Gebiete 2, 340-368 (1964; Zbl 0121.02406)] is proved. The investigations are based on an integration theory for vector measures due to Lebesgue-Pettis as well as Fubini type results concerning integration with respect to Markovian kernels, which are represented in an appendix.
      0 references

      Identifiers

      0 references
      0 references
      0 references
      0 references
      0 references