Module: tf_agents.bandits.policies.mixture_policy

A policy class that chooses from a set of policies to get the actions from.

This mixture policy takes a list of policies and will randomly choose one of them for every observation. The distribution is defined by the mixture_distribution.

Classes

class MixturePolicy: A policy that chooses from a set of policies to decide the action.

Other Members

MIXTURE_AGENT_ID = 'mixture_agent_id'
SUBPOLICY_INFO = 'subpolicy_info'

Except as otherwise noted, the content of this page is licensed under the Creative Commons Attribution 4.0 License, and code samples are licensed under the Apache 2.0 License. For details, see the Google Developers Site Policies. Java is a registered trademark of Oracle and/or its affiliates.

Last updated 2020-04-08.