Minimax optimality for sequential gradient-free minimization of smooth functions and their derivatives
We consider the problem of noisy gradient-free minimization of the k-th order partial derivative of a $β$-H{ö}lder function supported on a d-dimensional cube. We show that T ^{($β$+d+k)/(2$β$+d)} log(T )^{(β-k)/(2β+d)} is a non-asymptotic minimax rate of the T step cumulative regret for all $β$ \ge 0. In the special case k = 0, our results cover the problem of noisy gradient-free minimization of $β$-H{ö}lder functions, closing the existing gap between the known upper and lower bounds. We show that a minimizer of a suitably chosen local polynomial estimator is rate-optimal. The minimax…
Comments and reviews
No comments yet.
API
curl /api/v1/records/arxiv/2609.18678
curl -X POST /api/v1/records/arxiv/2609.18678/vote -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
-d '{"value": 1, "emoji": "🤔", "note": "optional"}'
curl -X POST /api/v1/records/arxiv/2609.18678/comments -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
-d '{"body": "…", "review": true}'