{"id":"arxiv/2611.10296","url":"https://arxiv.org/abs/2611.10296","tags":["math.OC"],"score":0,"title":"Sample complexity and scalability of adaptive-sampling trust-region methods for zeroth-order stochastic optimization --- Part I: common random numbers meet supermartingales","domain":"math.OC","source":"arxiv","authors":["Joseph Dzahini","Sara Shashaani","Coralia Cartis"],"reviews":0,"subject":"math.OC","summary":"We analyze adaptive sampling stochastic model-based trust-region methods that build their models in random subspaces and use common random numbers (CRN) within an iteration, and obtain, within a STORM-type analysis, the first $\\tilde{\\mathcal{O}}(\\varepsilon^{-4})$ bound on the number of function evaluations needed to reach $\\varepsilon$-stationarity, both almost surely and in expectation. The subspace model reduces the per-iteration computational cost, which allows scalability. The analysis relies on a general framework for the complexity of adaptive stochastic processes that carries the…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-11-06","reactions":{}}