Y
Yee Whye Teh
Researcher at University of Oxford
Publications - 351
Citations - 42930
Yee Whye Teh is an academic researcher from University of Oxford. The author has contributed to research in topics: Computer science & Inference. The author has an hindex of 68, co-authored 326 publications receiving 36155 citations. Previous affiliations of Yee Whye Teh include University of Toronto & University College London.
Papers
More filters
Proceedings Article
Particle Value Functions
Chris J. Maddison,Dieterich Lawson,George Tucker,Nicolas Heess,Arnaud Doucet,Andriy Mnih,Yee Whye Teh +6 more
TL;DR: In this article, the authors introduce the particle value function defined by a particle filter over the distributions of an agent's experience, which bounds the risk-sensitive one and illustrate the benefit of the policy gradients of this objective in Cliffworld.
Posted Content
The Mondrian Kernel
TL;DR: The Mondrian kernel as mentioned in this paper is a fast random feature approximation to the Laplace kernel, which is suitable for both batch and online learning, and admits a fast kernel-width selection procedure as the random features can be re-used efficiently for all kernel widths.
Bethe free energy and contrastive divergence approximations for undirected graphical models
Geoffrey E. Hinton,Yee Whye Teh +1 more
TL;DR: This thesis develops RBMrate, a model for discretized continuous-valued data and describes sparse and over-complete representations of data where the inference process is trivial since it is simply an EBM, and contributes a theory relating belief propagation and iterative scaling to the Bethe free energy approximations.
Posted Content
Fast MCMC sampling for Markov jump processes and extensions
Vinayak Rao,Yee Whye Teh +1 more
TL;DR: In this article, an auxiliary variable Gibbs sampler for continuous-time Markov chains is proposed. But it is based on the idea of uniformization, and does not involve approximations like time-discretization.
Proceedings ArticleDOI
Conformal Off-Policy Prediction in Contextual Bandits
TL;DR: This work proposes conformal off-policy prediction (COPP), which can output reliable predictive intervals for the outcome under a new target policy and provides theoretical and empirical guarantees about the utility of COPP compared with existing methods on synthetic and real-world data.