Search papers, labs, and topics across Lattice.
This paper establishes strong laws of large numbers (SLLNs) for locally Lipschitz functions within the Lipschitz pseudometric framework, addressing both topological and model-theoretic conditions. The findings extend previous work by identifying a wider array of functions that avoid failure phenomena previously documented, thereby enhancing the understanding of convergence properties in random function analysis. Key results include applications to uniform convergence of Clarke subdifferentials and finite-sample identification of solutions, providing a robust foundation for further theoretical exploration.
Broad classes of functions can now be analyzed without the failure phenomena previously observed, reshaping our understanding of convergence in random function contexts.
We prove strong laws of large numbers for locally Lipschitz functions in the Lipschitz pseudometric. Our results hold under either a topological or a model-theoretic condition, with the latter encompassing functions jointly definable in o-minimal structures but extending substantially beyond this class. Applications include uniform convergence of limiting and Clarke subdifferentials and finite-sample identification of solutions. Consequently, we identify broad classes of functions for which the failure phenomena revealed by our previous negative results [Tian and Royset, arXiv:2511.16568, 2025] do not occur.
