Computing fixed densities of a class of Frobenius-Perron operators

From MaRDI portal





Let \((X,{\mathcal X},\mu)\) be a \(\sigma\)-finite measure space, and let \(S:X\to X\) be a non-singular transformation, so that \(S\) is measurable and if \(A\in{\mathcal X}\) with \(\mu(A)= 0\), then \(\mu(S^{-1}(A))= 0\). The space consisting of integrable functions on \(X\) is denoted by \(L^1= L^1(X)\). The Frobenius-Perron operator \(P: L^1(X)\to L^1(X)\) associated with \(S\) is defined by \[ \int_A P(f) d\mu= \int_{S^{-1}} f d\mu. \] The set of density functions of the form \(f\) such that \(f\geq 0\) and \(\|f\|= \int_X|f|d\mu= 1\) is denoted by \(D\), and the operator \(M\) is said to be a Markov operator if \(f\geq 0\) implies \(M(f)\geq 0\) and \(\|M(f)\|=\|f\|\). The results of this paper relate to the existence of and construction of fixed densities of the form \(f^*\) such that \(P(f^*)= f^*\), for some cases in which \(X= \Omega\), a bounded closed region of Euclidean space \(\mathbb{R}^N\) and \(\mu= m\), the Lebesgue measure. In the first part of the paper it is shown that if \(A\) is a non-empty closed convex subset of \(L^\infty(\Omega)\cap D\), \(P(A)\subseteq A\), and \(\|P(f)\|_\infty\leq a\|f\|_\infty+ b\) \((f\in A)\), where \(a< 1\), \(b>0\), then there is a fixed density \(f^*\) of \(P\) in \(A\). In section 3 of the paper a fixed density \(f^*\) of \(P\) is defined in terms of a sequence \(\{Q_n\}\) of Markov finite-dimensional approximations of the identity.











This page was built for publication: Computing fixed densities of a class of Frobenius-Perron operators

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4264304)