On a metric space it suffices to test the variational criterion for relative entropy against bounded Lipschitz functions; consequently sublevel sets of the relative entropy are sequentially closed under weak convergence.
In the settings of The Real Numbers: Standing Notation and Background and Measure Spaces and the Lebesgue Integral: Standing Notation, the measure space of the latter being instantiated at each use by the measure space named, let be a metric space with nonempty and Borel -algebra , and let and be Borel measures on with . Finite relative entropy with respect to and are those of that definition on ; bounded measurable functions and the numbers are those of Relative Entropy on a Measurable Space: the Gibbs Inequality, the Variational Criterion and Formula, the Entropy Inequality, Small Sets and Data Processing for the measurable space . A function is called bounded Lipschitz if it is bounded and Lipschitz with a constant for and the absolute-value metric; by claim 1 below and Relative Entropy on a Measurable Space: the Gibbs Inequality, the Variational Criterion and Formula, the Entropy Inequality, Small Sets and Data Processing §functional, is a real number for every bounded Lipschitz . Weak convergence of Borel measures on is written .
1. (Lipschitz functions) Every bounded Lipschitz function is bounded measurable.
2. (Lipschitz criterion) Let be such that for every bounded Lipschitz . Then has finite relative entropy with respect to , and .
3. (Sequentially closed sublevel sets) Let and let be a sequence of Borel measures on with , each of finite relative entropy with respect to with , such that . Then has finite relative entropy with respect to , and .
Loading…
No relations recorded yet.