Anti Imitation-Based Policy Learning

Sebag, Michèle; Akrour, Riad; Mayeur, Basile; Schoenauer, Marc

doi:10.1007/978-3-319-46227-1_35

Cited by 2 publications

(1 citation statement)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Traditional IRL methods learn from (near) optimal demonstration. More recent approaches extend IRL to learn from other types of observations, e.g., a set of (nonnecessarily optimal) demonstrations rated by an expert [El Asri et al, 2016;Burchfield et al, 2016], bad demonstrations [Sebag et al, 2016] or pairwise comparisons [da Silva et al, 2006;Wirth and Neumann, 2015]. In the latter case, the interactive setting is investigated with a reliable expert [Chernova and Veloso, 2009] or unreliable one [Weng et al, 2013].…”

Section: Reward Learningmentioning

confidence: 99%

Reinforcement Learning

Buffet,

Pietquin,

Weng

2020

Preprint

View full text Add to dashboard Cite

Reinforcement learning (RL) is a general framework for adaptive control, which has proven to be efficient in many domains, e.g., board games, video games or autonomous vehicles. In such problems, an agent faces a sequential decisionmaking problem where, at every time step, it observes its state, performs an action, receives a reward and moves to a new state. An RL agent learns by trial and error a good policy (or controller) based on observations and numeric reward feedback on the previously performed action. In this chapter, we present the basic framework of RL and recall the two main families of approaches that have been developed to learn a good policy. The first one, which is value-based, consists in estimating the value of an optimal policy, value from which a policy can be recovered, while the other, called policy search, directly works in a policy space. Actor-critic methods can be seen as a policy search technique where the policy value that is learned guides the policy improvement. Besides, we give an overview of some extensions of the standard RL framework, notably when risk-averse behavior needs to be taken into account or when rewards are not available or not known.

show abstract

Section: Reward Learningmentioning

confidence: 99%