AI Chat AI Image Generator AI Video Text to Speech

A Second-Order Method for Stochastic Bandit Convex Optimisation

02/10/2023

∙

by Tor Lattimore, et al.

∙

∙

We introduce a simple and efficient algorithm for unconstrained zeroth-order stochastic convex bandits and prove its regret is at most (1 + r/d)[d^1.5√(n) + d^3] polylog(n, d, r) where n is the horizon, d the dimension and r is the radius of a known ball containing the minimiser of the loss.

Tor Lattimore
57 publications
András György
43 publications

page 1

page 2

page 3

page 4

research

∙ 05/31/2020

Improved Regret for Zeroth-Order Adversarial Bandit Convex Optimisation

We prove that the information-theoretic upper bound on the minimax regre...

0 Tor Lattimore, et al. ∙

research

∙ 02/01/2023

Bandit Convex Optimisation Revisited: FTRL Achieves Õ(t^1/2) Regret

We show that a kernel estimator using multiple function evaluations can ...

0 David Young, et al. ∙

research

∙ 03/10/2021

Linear Bandits on Uniformly Convex Sets

Linear bandit algorithms yield 𝒪̃(n√(T)) pseudo-regret bounds on compact...

0 Thomas Kerdreux, et al. ∙

research

∙ 02/10/2021

An Efficient Pessimistic-Optimistic Algorithm for Constrained Linear Bandits

This paper considers stochastic linear bandits with general constraints....

0 Xin Liu, et al. ∙

research

∙ 02/12/2022

Corralling a Larger Band of Bandits: A Case Study on Switching Regret for Linear Bandits

We consider the problem of combining and learning over a set of adversar...

0 Haipeng Luo, et al. ∙

research

∙ 06/01/2021

Minimax Regret for Bandit Convex Optimisation of Ridge Functions

We analyse adversarial bandit convex optimisation with an adversary that...

0 Tor Lattimore, et al. ∙

research

∙ 09/14/2016

Stochastic Heavy Ball

This paper deals with a natural stochastic optimization procedure derive...

0 Sébastien Gadat, et al. ∙