Module: tf_agents.bandits.policies.falcon_reward_prediction_policy

Stay organized with collections Save and categorize content based on your preferences.

Policy that samples actions based on the FALCON algorithm.

This policy implements an action sampling distribution based on the following paper: David Simchi-Levi and Yunzong Xu, "Bypassing the Monster: A Faster and Simpler Optimal Algorithm for Contextual Bandits under Realizability", Mathematics of Operations Research, 2021. https://arxiv.org/pdf/2003.12699.pdf

Classes

class FalconRewardPredictionPolicy: Policy that samples actions based on the FALCON algorithm.

Functions

get_number_of_trainable_elements(...): Gets the total # of elements in the network's trainable variables.