Learning from Demonstrations: Is It Worth Estimating a Reward Function?

33 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Hyperparameter Selection for Imitation Learning (2021, background): “Both approaches have their advantages and drawbacks (Piot et al., 2013).”
SQIL: Imitation Learning via Reinforcement Learning with Sparse Rewards (2019, methodology): “itude of the right-hand side of Equation 7, which is equivalent to minimizing jR q(s;a)j. The purpose of the penalty on jR q(s;a)jis two-fold: (1) it imposes a sparsity prior motivated by prior work [Piot et al., 2013], and (2) it incorporates information about the state transition dynamics into the”
Boosted and reward-regularized classification for apprenticeship learning (2014, background): “This approach has been theoretically studied in the finite horizon case [23, 18, 8] and in the infinite horizon case [14].”
Boosted Bellman Residual Minimization Handling Expert Demonstrations (2014, methodology): “The minimization of the criterion JE(Q) is known in the literature and used in the LfD paradigm [19,13,17].”
Primal Wasserstein Imitation Learning (2020, background): “recovering the environment’s reward) they are less sensitive to distributional shift (Pomerleau, 1991), they generalize to environments with different dynamics (Piot et al., 2013), and they can recover a near-optimal agent from suboptimal demonstrations (Brown et al.”
Score-based Inverse Reinforcement Learning (2016, methodology): “Here, we consider a special case of Garnets specified by three parameters: (NS , NA, NB).”