Create and explore imaginative 3D worlds with our new product Dreamfield (opens in a new tab)
Papers
Topics
Authors
Recent
Search
2000 character limit reached

Dual Averaging on Compactly-Supported Distributions And Application to No-Regret Learning on a Continuum

Published 29 Apr 2015 in cs.LG and math.OC | (1504.07720v1)

Abstract: We consider an online learning problem on a continuum. A decision maker is given a compact feasible set SS, and is faced with the following sequential problem: at iteration~tt, the decision maker chooses a distribution x(t)∈Δ(S)x^{(t)} \in \Delta(S), then a loss function ℓ(t):S→R+\ell^{(t)} : S \to \mathbb{R}_+ is revealed, and the decision maker incurs expected loss ⟨ℓ(t),x(t)⟩=Es∼x(t)ℓ(t)(s)\langle \ell^{(t)}, x^{(t)} \rangle = \mathbb{E}_{s \sim x^{(t)}} \ell^{(t)}(s). We view the problem as an online convex optimization problem on the space Δ(S)\Delta(S) of Lebesgue-continnuous distributions on SS. We prove a general regret bound for the Dual Averaging method on L2(S)L^2(S), then prove that dual averaging with ω\omega-potentials (a class of strongly convex regularizers) achieves sublinear regret when SS is uniformly fat (a condition weaker than convexity).

Authors (1)
Citations (2)

Summary

No one has generated a summary of this paper yet.

Paper to Video (Beta)

No one has generated a video about this paper yet.

Whiteboard

No one has generated a whiteboard explanation for this paper yet.

Open Problems

We haven't generated a list of open problems mentioned in this paper yet.

Continue Learning

We haven't generated follow-up questions for this paper yet.