Learning Graphical Models
A Background
In multi-scale RNNs (Chung et al., 2016), faster features predict when to hand control to the slower features. CHiVE (Kenter et al., 2019) is a temporally abstract hierarchy for speech synthesis.
A Proofs of Propositions
Deep directed generative models with energy-based probability estimation.
Posterior Refinement Improves Sample Efficiency in Bayesian Neural Networks
Monte Carlo (MC) integration is the de facto method for approximating the predictive distribution of Bayesian neural networks (BNNs).
Minimax Regret for Stochastic Shortest Path
We study the Stochastic Shortest Path (SSP) problem in which an agent has to reach a goal state in minimum total expected cost. In the learning formulation of the problem, the agent has no prior knowledge about the costs and dynamics of the model. She repeatedly interacts with the model for K episodes, and has to minimize her regret.
Finite-Sample Maximum Likelihood Estimation of Location
This problem is known as location parameter estimation in the statistics literature.
Variational Bayesian Reinforcement Learning with Regret Bounds
In reinforcement learning the Q-values summarize the expected future rewards that the agent will attain.
