A dual representation theorem on the conditional Orlicz space generated from a random normed module
Xia Zhang, Ke Qian, Ming Liu
TL;DR
The paper advances the theory of conditional Orlicz spaces on random normed modules by (i) introducing random Orlicz functions and defining the conditional Orlicz spaces $L_{\\mathcal{F}}^{\Phi}(E)$ and $H_{\\mathcal{F}}^{\Phi}(E)$, (ii) proving the denseness of the Orlicz heart in the RN module under the $(\varepsilon,\lambda)$-topology and (iii) establishing a dual representation linking the dual of the conditional Orlicz heart to the conditional Orlicz space built from the random conjugate space, with isometric isomorphism given by $[T f](x)=E[f(x)\mid \mathcal{F}]$ for $f\in L_{\\mathcal{F}}^{\Psi}(E^*)$. These results generalize and refine existing dualities in the literature, including special cases that recover classical $L^p$ and Orlicz dualities, and provide a robust framework for conditional risk measures in random environments.
Abstract
In this paper, we first introduce the notion of a random Orlicz function, and further present the conditional Orlicz space generated from a random normed module. Second, we prove the denseness of the Orlicz heart of a random normed module $E$ in $E$ with respect to the $(\varepsilon, λ)$-topology. Finally, based on the above work, we establish a dual representation theorem on the conditional Orlicz space generated from a random normed module, which extends and improves some known results.
