Learning agents for uncertain environments (extended abstract)

By Stuart J. Russell. 1998

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

Reinforcement learning in robotics: A survey (2013, background): “Russell (1998) brought the field to the attention of the machine learning community.”
Cooperative Inverse Reinforcement Learning (2016, background): “The ﬁeld of inverse reinforcement learning or IRL (Russell, 1998; Ng & Russell, 2000; Abbeel & Ng, 2004) is certainly relevant to the value alignment problem.”
An Algorithmic Perspective on Imitation Learning (2018, methodology): “…learning into directly replicating desired behavior (sometimes called behavioral cloning [Bain and Sammut, 1996]) and learning the hidden objectives of the desired behavior from demonstrations (called inverse optimal control [Kalman, 1964] or inverse reinforcement learning [Russell, 1998]).”
Model-Free Imitation Learning with Policy Optimization (2016, methodology): “An alternative and often more practical approach, called imitation learning , is to encode preferences and differentiate between desirable and undesirable outcomes using demonstrations provided by an expert (Pomerleau, 1991; Russell, 1998).”
Preference-Based Reinforcement Learning: A Preliminary Survey (2013, background): “IRL is about learning the reward function of the MDP from demonstrated solutions [13].”
A Survey of Inverse Reinforcement Learning: Challenges, Methods and Progress (2018, background): “[2] A. Ng, S. Russell, Algorithms for inverse reinforcement learning, Proceedings of the Seventeenth International Conference on Machine Learning 0 (2000) 663–670.”
Global Autoregressive Models for Data-Efficient Sequence Learning (2019, background): “We could say that while Training-2 addresses a question directly related to Reinforcement Learning, Training-1 addresses one related to Inverse Reinforcement Learning (Russell, 1998; Ng and Russell, 2000): it derives a reward from training evidence rather than imposing it externally.”
Distributional Reinforcement Learning for Energy-Based Sequential Models (2019, background): “8 If we think in the terms of the current paper, we can further observe that while Training-2 has direct connections to RL (exploiting a given reward to obtain a policy), Training-1 has some similarities to Inverse RL [19, 12]: deriving a reward from the training data, here purely inside a max-likel”
Learning from Demonstrations: Is It Worth Estimating a Reward Function? (2013, methodology): “These two methods are Apprenticeship Learning (AL) [1] and Inverse Reinforcement Learning (IRL) [7].”
Imitation Learning via Kernel Mean Embedding (2018, background): “A more principled approach is apprenticeship learning (AL), where the policy is sought that is guaranteed to perform at least as well as the expert (Russell 1998; Ng, Russell, and others 2000; Abbeel and Ng 2004).”
Integrating Behavior Cloning and Reinforcement Learning for Improved Performance in Sparse Reward Environments (2019, background): “other learning modalities such as inverse reinforcement learning (IRL), where it is desired to learn the reward function that the agent, or demonstrator, is optimizing to perform the task [25].”
A survey of robot learning from demonstration (2009, unknown): “This approach to obtaining a policy is in contrast to other techniques in which a policy is learned from experience, for example building a policy based on data acquired through exploration , as in Reinforcement Learning [1].”
Inverse Reinforcement Learning (2010, methodology): “Although reinforcement learning methods are largely used to fully observable MDPs where the sensor input is sufﬁcient to identify the states, RL can also be applied on partially observable MDPs (POMDP)[5].”
Bayesian Inverse Reinforcement Learning (2007, background): “The Inverse Reinforcement Learning (IRL) problem is defined in [Russell, 1998] as follows: Determine The reward function that an agent is optimizing.”
A Bayesian Approach to Generative Adversarial Imitation Learning (2018, background): “Inverse reinforcement learning (IRL) [Russell, 1998, Ng and Russell, 2000, Ziebart et al., 2008] is a more modern approach, where the objective is to learn the underlying unknown cost function that makes the expert optimal.”
What Would Jiminy Cricket Do? Towards Agents That Behave Morally (2021, background): “Inverse reinforcement learning estimates reward functions by observing optimal agent behavior (Russell, 1998).”