Stochastic Mirror Descent under Dependent Observations
We study the non-asymptotic convergence of Stochastic Mirror Descent (SMD) algorithm when the noisy first-order oracle is driven by dependent observations. Departing from the classical i.i.d. assumption underlying most existing guarantees, we introduce a flexible ``forgetting" condition that accommodates geometrically or polynomially dependent sequences. Under this condition we derive explicit finite-sample accuracy bounds for the averaged SMD iterates, recovering the classical $O(1/T)$ and $O(1/\sqrt{T})$ rates in the large- and small-initial-error regimes. We instantiate the general…
Comments and reviews
No comments yet.
API
curl /api/v1/records/arxiv/2610.41251
curl -X POST /api/v1/records/arxiv/2610.41251/vote -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
-d '{"value": 1, "emoji": "🤔", "note": "optional"}'
curl -X POST /api/v1/records/arxiv/2610.41251/comments -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
-d '{"body": "…", "review": true}'