paper-with-me

Papers

Generalised Mixability, Constant Regret, and Bayesian Updating

2014-03-10 · Mark D. Reid, Rafael M. Frongillo, Robert C. Williamson

Mixability of a loss is known to characterise when constant regret bounds are achievable in games of prediction with expert advice through the use of Vovk's aggregating algorithm. We provide a new interpretation of mixability via convex analysis that highlights the role of the Kullback-Leibler divergence in its definition. This naturally generalises to what we call $\Phi$-mixability where the Bregman divergence $D_\Phi$ replaces the KL divergence. We prove that losses that are $\Phi$-mixable also enjoy constant regret bounds via a generalised aggregating algorithm that is similar to mirror descent.

📄 PDF Abstract BibTeX arXiv:1403.2433

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Constant Regret, Generalized Mixability, and Mirror Descent

2018-02-20 · NeurIPS 2018 12 · Zakaria Mhammedi, Robert C. Williamson

We consider the setting of prediction with expert advice; a learner makes predictions by aggregating those of a group of experts. Under this setting, and for the right choice of loss function and "mixing" algorithm, it i…

Open-Ended Question Answering

Generalized Mixability via Entropic Duality

2014-06-24 · Mark D. Reid, Rafael M. Frongillo, Robert C. Williamson, Nishant Mehta

Mixability is a property of a loss which characterizes when fast convergence is possible in the game of prediction with expert advice. We show that a key property of mixability generalizes, and the exp and log operations…

Mixability in Statistical Learning

2012-12-01 · NeurIPS 2012 12 · Tim V. Erven, Peter Grünwald, Mark D. Reid, Robert C. Williamson

Statistical learning and sequential prediction are two different but related formalisms to study the quality of predictions. Mapping out their relations and transferring ideas is an active area of investigation. We provi…

Bayesian InferencePrediction

An Information-Theoretic Approach to Minimax Regret in Partial Monitoring

2019-02-01 · Tor Lattimore, Csaba Szepesvari

We prove a new minimax theorem connecting the worst-case Bayesian regret and minimax regret under partial monitoring with no assumptions on the space of signals or decisions of the adversary. We then generalise the infor…

Exploiting the Surrogate Gap in Online Multiclass Classification

2020-07-24 · NeurIPS 2020 12 · Dirk van der Hoeven

We present Gaptron, a randomized first-order algorithm for online multiclass classification. In the full information setting we show expected mistake bounds with respect to the logistic loss, hinge loss, and the smooth h…

ClassificationGeneral Classification