Module: tf_agents.bandits.agents.neural_falcon_agent

A neural network based agent that implements Falcon exploration.

The policy samples actions with the action distribution proposed in the FALCON 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 NeuralFalconAgent: A neural network based agent implementing the Falcon sampling strategy.