We present a tutorial on Bayesian optimization, a method of finding the
maximum of expensive cost functions. Bayesian optimization employs the Bayesian
technique of setting a prior over the objective function and combining it with
evidence to get a posterior function. This permits a utility-based selection of
the next observation to make on the objective function, which must take into
account both exploration (sampling from areas of high uncertainty) and
exploitation (sampling areas likely to offer improvement over the current best
observation). We also present two detailed extensions of Bayesian optimization,
with experiments---active user modelling with preferences, and hierarchical
reinforcement learning---and a discussion of the pros and cons of Bayesian
optimization based on our experiences.
Description
A Tutorial on Bayesian Optimization of Expensive Cost Functions, with
Application to Active User Modeling and Hierarchical Reinforcement Learning
%0 Generic
%1 brochu2010tutorial
%A Brochu, Eric
%A Cora, Vlad M.
%A de Freitas, Nando
%D 2010
%K bayesian
%T A Tutorial on Bayesian Optimization of Expensive Cost Functions, with
Application to Active User Modeling and Hierarchical Reinforcement Learning
%U http://arxiv.org/abs/1012.2599
%X We present a tutorial on Bayesian optimization, a method of finding the
maximum of expensive cost functions. Bayesian optimization employs the Bayesian
technique of setting a prior over the objective function and combining it with
evidence to get a posterior function. This permits a utility-based selection of
the next observation to make on the objective function, which must take into
account both exploration (sampling from areas of high uncertainty) and
exploitation (sampling areas likely to offer improvement over the current best
observation). We also present two detailed extensions of Bayesian optimization,
with experiments---active user modelling with preferences, and hierarchical
reinforcement learning---and a discussion of the pros and cons of Bayesian
optimization based on our experiences.
@misc{brochu2010tutorial,
abstract = {We present a tutorial on Bayesian optimization, a method of finding the
maximum of expensive cost functions. Bayesian optimization employs the Bayesian
technique of setting a prior over the objective function and combining it with
evidence to get a posterior function. This permits a utility-based selection of
the next observation to make on the objective function, which must take into
account both exploration (sampling from areas of high uncertainty) and
exploitation (sampling areas likely to offer improvement over the current best
observation). We also present two detailed extensions of Bayesian optimization,
with experiments---active user modelling with preferences, and hierarchical
reinforcement learning---and a discussion of the pros and cons of Bayesian
optimization based on our experiences.},
added-at = {2016-08-25T17:58:42.000+0200},
author = {Brochu, Eric and Cora, Vlad M. and de Freitas, Nando},
biburl = {https://www.bibsonomy.org/bibtex/29afbd0a0feca633dd063f7ce55eed526/stdiff},
description = {A Tutorial on Bayesian Optimization of Expensive Cost Functions, with
Application to Active User Modeling and Hierarchical Reinforcement Learning},
interhash = {3f3272d1ac9e4d1e667015a512c3a8e1},
intrahash = {9afbd0a0feca633dd063f7ce55eed526},
keywords = {bayesian},
note = {cite arxiv:1012.2599},
timestamp = {2016-08-25T17:58:42.000+0200},
title = {A Tutorial on Bayesian Optimization of Expensive Cost Functions, with
Application to Active User Modeling and Hierarchical Reinforcement Learning},
url = {http://arxiv.org/abs/1012.2599},
year = 2010
}