Hybrid RL learns from expert state sequences without full action data.
problem Learning from expert state sequences without full action data.
method Tensor-based model to infer unobserved actions; hybrid RL objective.
result Hybrid RL outperforms pure RL and tensor-based action inference.
Survey of methods for learning from observation without requiring expert actions.
problem Lack of access to expert actions in imitation learning.
method Survey and classification of state-only imitation learning (SOIL) methods.
result Identification of open problems and future research directions.
A new method for multi-expert learning-to-defer avoids optimization issues.
problem Optimization issues in multi-expert learning-to-defer systems.
method A decoupled surrogate with a softmax classifier head and independent sigmoid heads per expert.
result First multi-expert L2D guarantee with a constant not growing with the expert pool.
Paper tackles imitation learning with sparse rewards and heterogeneous actions.
problem Challenges of imitation learning with sparse rewards and different actions.
method Proposes a method that balances imitation and reinforcement learning objectives.
result Agent efficiently leverages sparse rewards and learns from different actions.
Bayesian algorithms improve online learning with adversaries over infinite action spaces.
problem Online learning with adversaries over infinite action spaces.
method Developed a Thompson sampling algorithm for online learning with an adversary's prior over the space of actions.
result Thompson sampling over a Gaussian process prior achieves a rate of O ( β T d log ( 1 + d λ β ) ) O(β\sqrt{Td\log(1+\sqrt{d}\fracλβ)}) O ( β T d log ( 1 + d β λ ) ) against a β β β -bounded λ λ λ -Lipschitz adversary. Continuous control imitation learning fails if expert actions are smooth.
problem Continuous control imitation learning fails if expert actions are smooth.
method Study of imitation learning in discrete-time, continuous state-and-action control systems.
result Any smooth, deterministic imitator policy suffers exponentially larger error than the expert.
Reinforcement learning methods require careful design involving a reward function to obtain the desired action policy for a given task. In the absence of hand-crafted reward functions, prior work on the topic has proposed several methods for reward estimation by using expert state trajectories and action pairs. However…
The paper proposes a method to explain expert decisions by modeling preferences with 'what if' outcomes.
problem Interpreting and auditing decision-making policies in institutions.
method Integrating counterfactual reasoning into batch inverse reinforcement learning.
result The method effectively recovers accurate and interpretable descriptions of expert behavior.
Paper proposes interpretable RL policies from a mixture of experts.
problem Making RL policies transparent and understandable in real-world applications.
method Policy iteration scheme with interpretable experts and prototypical states.
result Proposed algorithm learns policies comparable to neural networks but more interpretable.
Meta-algorithm optimizes nonstochastic bandits with infinitely many experts.
problem Maximizing reward by choosing actions sequentially from a set of experts.
method Proposed a variant of Exp4.P for infinitely many experts and a meta-algorithm.
result Proved high-probability upper bound of i l d e O ( i ∗ K + K T ) ilde{\mathcal{O}} \big( i^*K + \sqrt{KT} \big) i l d e O ( i ∗ K + K T ) on regret. Current imitation learning techniques are too restrictive because they require the agent and expert to share the same action space. However, oftentimes agents that act differently from the expert can solve the task just as good. For example, a person lifting a box can be imitated by a ceiling mounted robot or a desktop…
Paper tackles covariate shift in offline IL using less proficient behavior data.
problem Mitigating covariate shift in Imitation Learning with limited data coverage.
method Model-based IL from Offline Data (MILO) framework.
result MILO can combat covariate shift even with sub-optimal behavior policy data.
Paper proposes a method to learn and exceed expert demonstrations in unknown reward environments.
problem Learning to outperform expert demonstrations in unknown reward environments.
method A novel concurrent reward and action policy learning approach with a stereo utility definition.
result The proposed method can outperform expert demonstrations in various environments.
New method recovers diverse policies from expert data using state-action pair weighting.
problem Recovering diverse policies from expert trajectories.
method Pointwise mutual information weighted behavioral cloning.
result Effective in focusing on state-action pairs most representative of the style.
Efficient algorithms for online learning with changing action sets, achieving no-approximate-regret guarantees.
problem Online learning with sleeping experts/bandits, where only a subset of actions are available each time.
method Developed computationally efficient algorithms providing no-approximate-regret guarantees for the general problem and better approximation ratios for special cases.
result Achieved no-approximate-regret guarantees for the general sleeping expert/bandit problems and better approximation ratios for specific cases.
Improved regret bounds for bandits with fixed expert advice using information theory.
problem Optimizing regret in bandit problems with fixed expert distributions.
method Information-theoretic analysis and KL-divergence measures.
result First regret bounds for EXP4 that can get arbitrarily close to zero under certain conditions.
In this work we describe a novel deep reinforcement learning architecture that allows multiple actions to be selected at every time-step in an efficient manner. Multi-action policies allow complex behaviours to be learnt that would otherwise be hard to achieve when using single action selection techniques. We use both …
Improved regret bounds for bandits with expert advice.
problem Optimizing decision-making in environments with expert advice.
method Proved lower and upper bounds for regret in restricted and standard feedback models.
result Proved a new upper bound of order K T ln ( N / K ) \sqrt{K T \ln(N/K)} K T ln ( N / K ) for the worst-case regret, matching a previously known lower bound. The current dominant paradigm for imitation learning relies on strong supervision of expert actions to learn both 'what' and 'how' to imitate. We pursue an alternative paradigm wherein an agent first explores the world without any expert supervision and then distills its experience into a goal-conditioned skill policy …
FinFlowRL learns from experts to optimize financial control in changing markets.
problem Traditional finance control methods fail in real-world, non-stationary markets.
method Imitation-Reinforcement Learning framework that pretrains on expert strategies and finetunes in noise space.
result Consistently outperforms individually optimized experts across diverse market conditions.
A new algorithm trains experts to safely guide agents in partially observed environments.
problem Existing imitation learning methods for POMDPs can lead to sub-optimal or unsafe policies.
method Derive an objective to encourage the expert to maximize the agent's reward, then use it to train both expert and agent.
result The algorithm produces an expert policy that the agent can safely imitate, outperforming fixed expert policies.
We consider an adversarial online learning setting where a decision maker can choose an action in every stage of the game. In addition to observing the reward of the chosen action, the decision maker gets side observations on the reward he would have obtained had he chosen some of the other actions. The observation str…
The goal of task transfer in reinforcement learning is migrating the action policy of an agent to the target task from the source task. Given their successes on robotic action planning, current methods mostly rely on two requirements: exactly-relevant expert demonstrations or the explicitly-coded cost function on targe…
The paper shows how expert knowledge can improve treatment effect estimation.
problem Lack of leveraging expert knowledge in treatment effect estimation.
method Formally defining two types of expertise (predictive and prognostic) and demonstrating their influence on treatment effect estimation methods.
result Expertise type significantly influences treatment effect estimation methods, and can be predicted from a dataset.
Improves reward bounds for prediction with expert advice using abstention.
problem Prediction with expert advice under bandit feedback with abstention.
method CBA algorithm exploiting abstention to improve reward bounds.
result Achieved significant improvement in reward bounds for general confidence-rated predictors.
MobILE learns from expert demonstrations without access to actions, achieving strong performance guarantees.
problem Imitation Learning from Observations alone (ILFO) where actions are not observed.
method MobILE integrates optimism in the face of uncertainty into distribution matching IL framework, trading off exploration and imitation.
result MobILE provides strong performance guarantees for certain MDP dynamics and demonstrates efficacy on OpenAI Gym tasks.
FinFlowRL combines imitation and reinforcement learning for better financial control.
problem Traditional stochastic control methods fail in real-world finance due to changing market conditions.
method FinFlowRL uses imitation learning to pretrain an adaptive meta policy, then finetunes it with reinforcement learning.
result FinFlowRL consistently outperforms individual strategies across various market conditions.
Near-optimal per-action regret bounds for sleeping bandits are derived.
problem Optimizing performance in sleeping bandits where arms and losses are chosen by an adversary.
method Directly minimizing per-action regret using generalized versions of EXP3, EXP3-IX, and FTRL with Tsallis entropy.
result Near-optimal bounds of order O ( T A ln K ) O(\sqrt{TA\ln{K}}) O ( T A ln K ) and O ( T A K ) O(\sqrt{T\sqrt{AK}}) O ( T A K ) are obtained. Interactive IL algorithm queries noisy expert for feedback, reducing sample size.
problem Imitation learning with noisy expert feedback requires many samples.
method Selective sampling algorithm with online regression oracle.
result Interactive IL algorithm reduces query complexity and improves performance.
Improved sample efficiency in reinforcement learning with action guidance.
problem Sample inefficiency in deep reinforcement learning with sparse, delayed, and deceptive rewards.
method Integrating a non-expert demonstrator (e.g., MCTS) into asynchronous distributed deep reinforcement learning.
result Our methods learn faster and converge to better policies on a game.
LEARN-SAM improves RL from sub-optimal demonstrations by localizing expert policies and selectively using demonstrations.
problem Improving RL from sub-optimal or sparse demonstrations.
method Local Ensemble and Reparameterization with Split and Merge of expert policies (LEARN-SAM).
result LEARN-SAM boosts learning speed and accuracy by selectively using demonstrations.
PWIL learns agent behavior from expert using Wasserstein distance.
problem Learning agent behavior from expert demonstrations efficiently.
method Primal Wasserstein Imitation Learning (PWIL) with offline reward function.
result PWIL achieves expert behavior on MuJoCo tasks with minimal interactions.
A new method for learning to defer decisions with expert advice improves over standard methods.
problem Learning to defer decisions with expert advice in systems where expert information can be modified after selection.
method An augmented surrogate that operates on the composite expert-advice action space, providing consistency guarantees and excess-risk bounds.
result The method improves over standard Learning-to-Defer and adapts its advice acquisition behavior to the cost regime.
LEAQI learns to query an expert less often for noisy guidance.
problem Active imitation learning with noisy guidance to reduce expert query complexity.
method Developed LEAQI, a difference classifier that predicts expert disagreement with noisy heuristic.
result Significantly fewer queries to the expert with comparable or better accuracy.
A new framework uses deep RL to aggregate expert advice for better portfolio management.
problem Improving portfolio management through expert advice and deep reinforcement learning.
method Convolutional networks for signal aggregation and historical price data, Proximal Policy Optimization algorithm.
result Our framework can achieve 90% of the best expert's profit on average.
State-only imitation learning improves dexterous manipulation learning from videos.
problem High sample complexity in complex domains like dexterous manipulation.
method Train an inverse dynamics model to predict actions from states and train the policy jointly.
result Performs on par with state-action approaches and outperforms RL alone.
The paper tackles efficient online learning by achieving minimal regret with respect to the best expert.
problem Achieving minimal regret in online learning problems where the goal is to match the lowest regret of K experts.
method A lazy form of the online subgradient algorithm is used to achieve minimal regret in 'easy' regimes.
result Minimal regret strategies exist for some 'hard' regimes, and the algorithm retains an O ( n ) O(\sqrt{n}) O ( n ) worst-case regret guarantee. Develops statistical framework for resolving reward function ambiguity in inverse reinforcement learning.
problem Non-uniqueness of reward functions in inverse reinforcement learning.
method Entropy regularization combined with least-squares reconstruction of the reward from the soft Bellman residual.
result Least-squares reward function is unique and consistent with the expert policy.
AI system narrows human decision options for better outcomes.
problem Improving human decision-making in sequential tasks.
method Developed a decision support system using a pre-trained AI agent to limit human action choices.
result Participants outperformed AI and solo play in a wildfire mitigation game.
We consider apprenticeship learning, i.e., having an agent learn a task by observing an expert demonstrating the task in a partially observable environment when the model of the environment is uncertain. This setting is useful in applications where the explicit modeling of the environment is difficult, such as a dialog…
Framework uses expert intervention to solve long-horizon reinforcement learning tasks.
problem Long horizon robot learning tasks with sparse rewards.
method Option templates and expert intervention to enable high-level task understanding.
result Framework outperforms state-of-the-art approaches by two orders of magnitude.
An assistant learns to mediate decisions between humans and experts, balancing risk and learning.
problem Learning to mediate decisions between imperfect humans and expert knowledge.
method Formalizes online decision mediation, proposes a policy to balance immediate loss and future generalization.
result Consistent gains over benchmarks in decision-making performance.
A new method for ILO with transition model disparity using an intermediary policy.
problem Learning tasks from expert observations with different transition dynamics.
method Training an intermediary policy to match the state transitions of the expert dataset.
result Our method outperforms existing ILO approaches with transition model mismatch.
New method aligns states for better imitation learning.
problem Different dynamics models between imitator and expert.
method State alignment-based reinforcement learning.
result Superior performance on various imitation learning settings.
GWIL uses Gromov-Wasserstein distance to align expert and imitation agent states.
problem Cross-domain imitation learning challenges due to different system dimensions and stationary distributions.
method Gromov-Wasserstein Imitation Learning (GWIL) using Gromov-Wasserstein distance.
result GWIL effectively aligns expert and imitation agent states in various continuous control domains.
Proposes hybrid reinforcement learning for both discrete and continuous control problems.
problem Real-world control problems involving both discrete and continuous decision variables.
method Solves hybrid problems by optimizing for discrete and continuous actions simultaneously.
result Efficiently solves hybrid reinforcement learning problems and improves upon expert heuristics.
We provide the first algorithm for online bandit linear optimization whose regret after T rounds is of order sqrt{Td ln N} on any finite class X of N actions in d dimensions, and of order d*sqrt{T} (up to log factors) when X is infinite. These bounds are not improvable in general. The basic idea utilizes tools from con…
XLabel tool reduces medical experts' workload by 40% and explains its decisions.
problem Efficiently labeling large electronic health records.
method Visual-interactive tool using Explainable Boosting Machine (EBM) for classification and explanation.
result EBM achieves high accuracy and explainability, even with mislabeled data.