Black-box policy search with probabilistic programs
In this work we show how to represent policies as programs: that is, as stochastic simulators with tunable parameters. To learn the parameters of such policies we develop connections between black box variational inference and existing policy search approaches. We then explain how such learning can...
Main Authors: | Van De Meent, J, Paige, B, Tolpin, D, Wood, F |
---|---|
格式: | Conference item |
出版: |
Journal of Machine Learning Research
2016
|
相似書籍
-
Output-sensitive Adaptive Metropolis-Hastings for probabilistic programs
由: Tolpin, D, et al.
出版: (2015) -
Maximum a posteriori estimation by search in probabilistic programs
由: Tolpin, D, et al.
出版: (2015) -
Bayesian Optimization for Probabilistic Programs
由: Rainforth, T, et al.
出版: (2016) -
Particle Gibbs with Ancestor Sampling for Probabilistic Programs
由: van de Meent, J, et al.
出版: (2015) -
Particle Gibbs with ancestor sampling for probabilistic programs
由: Meent, J, et al.
出版: (2015)