Apprenticeship Learning About Multiple Intentions

By Monica Babes-Vroman, Vukosi Marivate, Kaushik Subramanian, Michael L. Littman. 2011

197 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.

Citations in the corpus

The MAGICAL Benchmark for Robust Imitation (2020, neutral): “However, it’s worth noting that unlike, say, HAMDPs [16] or CIRL games [21], IL cannot request clarification from a demonstrator if the supplied demonstrations are ambiguous, which limits its ability to learn the right objective in general.”
A Survey of Reinforcement Learning Informed by Natural Language (2019, methodology): “…available, the reward function can be learned using standard IRL methods like MaxEnt IRL [Ziebart et al. , 2008] as in [Fu et al. , 2019], or maximum likelihood IRL [Babes et al. , 2011] as in [MacGlashan et al. , 2015], who also learn a joint generative model of rewards, behaviour, and language.”
Maximum Entropy Deep Inverse Reinforcement Learning (2015, methodology): “Furthermore, other methods for optimising demonstration data likelihood such as given by (Babes et al., 2011) will be evaluated.”
Imitation Learning as f-Divergence Minimization (2019, background): “Note that [43] also independently arrives at such connections be-tween f-divergence and IL. 3 We particularly focus on multi-modal expert demonstrations which has generally been treated by clustering data and learning on each cluster [44, 45].”
A New Softmax Operator for Reinforcement Learning (2016, methodology): “Algorithms for inverse reinforcement learning (IRL), the problem of inferring reward functions from observed behavior (Ng & Russell, 2000), frequently use a Boltzmann operator to avoid assigning zero probability to non-optimal actions and hence assessing an observed sequence as impossible.”
OptionGAN: Learning Joint Reward-Policy Options using Generative Adversarial Inverse Reinforcement Learning (2017, background): “To this end, several works investigate decomposing the underlying reward functions of expert demonstrations and environments in both IRL and RL (Krishnan et al. 2016; Sermanet, Xu, and Levine 2016; Choi and eung Kim 2012; Babes et al. 2011; van Seijen et al. 2017).”
Grounding English Commands to Reward Functions (2015, methodology): “Following maximum likelihood inverse reinforcement learning [1], we use a Boltzmann distribution over the optimal Qvalues as the noisy policy.”
Imitating Latent Policies from Observation (2018, background): “Other works have aimed to cluster demonstrations, but these approaches have traditionally segmented different types of trajectories, which represent distinct preferences, rather than next-state predictions (Hausman et al., 2017; Babes et al., 2011).”
Bayesian Multitask Inverse Reinforcement Learning (2011, methodology): “Our approach can be seen as a generalisation of [3] to the dynamic setting of inverse reinforcement learning; of [2] to full Bayesian estimation; and of [20] to multiple tasks.”
A survey of inverse reinforcement learning techniques (2012, methodology): “Using a gradient ascent approach, IRL is also extended to a multiple intention setting (Babes et al., 2010).”
A Survey of Inverse Reinforcement Learning: Challenges, Methods and Progress (2018, methodology): “rturbation. For example, mlirl allows tuning of its model parameter in Eq. 10 to allow more randomness into the learned policy ˇ Ewhen the demonstrated behavior is expected to be noisy and suboptimal [45]. On the other hand, methods such as mmp and learch introduce slack variables in their optimizat”
Reinforcement learning improves behaviour from evaluative feedback (2015, background): “The problem of inverse reinforcement learning 45,46 addresses the challenge of creating appropriate reward functions given the availability of behavioural traces from an expert executing the target behaviour.”
Learning a Prior over Intent via Meta-Inverse Reinforcement Learning (2018, background): “Other approaches have clustered demonstrations while simultaneously inferring reward functions for each cluster (Babeş-Vroman et al., 2011) or introduced regularization between rewards to a common “shared reward” (Li & Burdick, 2017).”
Multi-Modal Imitation Learning from Unstructured Demonstrations using Generative Adversarial Nets (2017, methodology): “Examples from this ﬁeld include [9] and [4].”
Showing versus doing: Teaching by demonstration (2016, methodology): “For instance, Maximum Likelihood IRL (MLIRL) is a state-of-the-art IRL algorithm for inferring feature-based reward functions [4, 12].”
An Alternative Softmax Operator for Reinforcement Learning (2016, methodology): “Algorithms for inverse reinforcement learning (IRL), the problem of inferring reward functions from observed behavior (Ng & Russell, 2000), frequently use a Boltzmann operator to avoid assigning zero probability to non-optimal actions and hence assessing an observed sequence as impossible.”