Develops a support-aware framework for reserve-policy selection in advertising markets.
problem Log-based reserve-price evaluation risks weak support and subgroup harm.
method Support-aware offline decision framework converting logged evidence into certified policies.
result Preserves the best gate-passing policy while eliminating only policies with certified regret.
CASP selects reliable policies for two-stage recommender systems by considering both value and support.
problem The selection of a generator in two-stage recommender systems affects both the policy value and the data support used to estimate it.
method CASP combines doubly robust value estimation with a support-burden penalty.
result CASP selects lower-burden policies when estimated value and support credibility are in tension.
Contextual bandit methods fail with deficient support data.
problem Learning from support-deficient data in contextual bandits.
method Three approaches to IPS-based learning: action space restriction, reward extrapolation, and policy space restriction.
result Systematic analysis and empirical evaluation of approaches to IPS-based learning.
A new method combines online and offline learning to tackle contextual bandits with missing action support.
problem Learning optimal policies with logged data when the logging policy has deficient support.
method Hybrid approach using online exploration to exploit supported actions and offline learning to avoid unnecessary explorations.
result Determines an optimal policy with theoretical guarantees using minimal online explorations.
A new framework estimates expert policy support to create a reward function for imitation learning.
problem Imitation learning from expert trajectories without reinforcement signals.
method Estimating the support of the expert policy to compute a fixed reward function.
result Comparable or better performance than state-of-the-art methods on discrete and continuous domains.
Extends OPE to evaluate policies using diverse logging data.
problem Evaluate policies using log data from different policies.
method Develops an OPE method for various logging policies.
result Method's predictions converge to true performance as sample size increases.
CQL (ReDS) learns from varied driving behaviors, improving offline RL performance.
problem Learning from datasets with non-uniform variability in behavior policies.
method Reweighting the data distribution to allow per-state flexibility in following the behavior policy.
result CQL (ReDS) improves performance in various offline RL tasks.
Paper addresses OPE for dependent bandit samples using MDS and batch updates.
problem Evaluating policies from non-i.i.d. historical data in contextual bandits.
method Constructs an MDS-based estimator for dependent samples, solves batch update and deficient support issues.
result Derives an asymptotically normal estimator for evaluation policy value.
Monotonic policy improvement and off-policy learning are two main desirable properties for reinforcement learning algorithms. In this paper, by lower bounding the performance difference of two policies, we show that the monotonic policy improvement is guaranteed from on- and off-policy mixture samples. An optimization …
Lapse-supported life insurance exacerbates adverse selection risks.
problem Lapse-supported life insurance increases adverse selection costs.
method Modeling 'Term to 100' contracts and analyzing three methods of managing lapse surplus.
result Adverse selection losses can be almost unlimited under certain conditions.
DOLCE improves off-policy evaluation and learning by decomposing effects.
problem Bias in off-policy evaluation and learning due to policy mismatch.
method Uses lagged contexts and a moment-based training procedure to decompose and cancel bias.
result DOLCE achieves substantial improvements in off-policy evaluation and learning.
We present a new approach to the problems of evaluating and learning personalized decision policies from observational data of past contexts, decisions, and outcomes. Only the outcome of the enacted decision is available and the historical policy is unknown. These problems arise in personalized medicine using electroni…
SAIL improves AIL by weighting adversarial rewards with support estimation.
problem Training instability and reward bias in AIL.
method Support-weighted Adversarial Imitation Learning (SAIL) extends AIL with support estimation to improve reinforcement signals.
result SAIL achieves better performance and stability on benchmark tasks.
Develops a new RL algorithm for medical treatment regimes.
problem Optimal dose determination in continuous action environments.
method Quasi-optimal learning algorithm for near-optimal actions.
result Guaranteed convergence and effectiveness in real applications.
Paper analyzes IRL problem and provides sample complexity analysis.
problem Finding a reward function for a given optimal policy.
method Geometric analysis and L1-regularized SVM formulation.
result Sample complexity of O(n^2 log(nk)) for recovering reward function.
Improves reinforcement learning stability and efficiency.
problem Combining stability and efficiency in reinforcement learning.
method Combines on-policy stability with off-policy sample reuse.
result Demonstrates improved performance in both theory and practice.
Managers of US National Forests must decide what policy to apply for dealing with lightning-caused wildfires. Conflicts among stakeholders (e.g., timber companies, home owners, and wildlife biologists) have often led to spirited political debates and even violent eco-terrorism. One way to transform these conflicts into…
New RL algorithms improve control tasks with data reuse.
problem Real-world control requires performance guarantees and data efficiency.
method Generalized Policy Improvement combining on-policy guarantees and sample reuse.
result Extensive experimental analysis shows benefits of new algorithms.
TREK uses distillation to help students solve hard problems.
problem Stalled progress on hard prompts when current policy lacks useful reasoning trajectories.
method TREK combines distillation and reinforcement learning to expand student support.
result TREK significantly improves student performance on mathematical reasoning and agentic tasks.
In some reinforcement learning problems an agent may be provided with a set of input policies, perhaps learned from prior experience or provided by advisors. We present a reinforcement learning with policy advice (RLPA) algorithm which leverages this input set and learns to use the best policy in the set for the reinfo…
SafePILCO is a Python tool for safe reinforcement learning.
problem Safe and efficient policy synthesis in reinforcement learning.
method Extends PILCO algorithm with safety features, implemented in Python.
result Safe and data-efficient policy synthesis achieved.
Deep reinforcement-learning methods have achieved remarkable performance on challenging control tasks. Observations of the resulting behavior give the impression that the agent has constructed a generalized representation that supports insightful action decisions. We re-examine what is meant by generalization in RL, an…
We consider the relationship between economic activity and intervention, including monetary and fiscal policy, using a universal dynamic framework. Central bank policies are designed for growth without excess inflation. However, unemployment, investment, consumption, and inflation are interlinked. Understanding dynamic…
This work improves policy-based training by proposing an evaluation balance objective for GFlowNets.
problem Reliable estimation of policy divergence under directed acyclic graphs remains challenging.
method Proposes an evaluation balance objective over partial episodes to measure policy divergence and improve policy-based training reliability.
result Evaluation balance strengthens policy-based training reliability and broadens its flexibility.
Paper improves safe policy improvement with estimated baseline policy.
problem Unreliable batch Reinforcement Learning algorithms in real-world applications.
method Apply SPIBB algorithms with an estimated baseline policy.
result Safe policy improvement guarantees over true baseline without direct access.
The paper examines how macroeconomic control tools lost effectiveness, leading to a 'dark ages' period.
problem Loss of effectiveness of control tools in macroeconomic stabilization policy.
method Historical analysis of macroeconomic stabilization policy from 1948 to 1993.
result The overstatement of the Lucas critique and Kydland and Prescott's time-inconsistency led to a period of ineffective stabilization policy.
MOPO optimizes offline RL by penalizing dynamics uncertainty.
problem Learning policies from offline data with distributional shift.
method Modify model-based RL to avoid distributional shift.
result MOPO outperforms model-free and standard model-based RL.
Optimal learning for parametric prophet inequalities with exponential-type distributions
problem Learning in prophet inequalities with unknown parameters
method Confidence-based dynamic-programming policy
result Achieves optimal asymptotic competitive ratio using online observations
Bayesian approach models policy distribution for faster exploration and transfer learning.
problem Challenges in exploration and adaptation to new tasks in transfer learning.
method Modeling a distribution over policies in a Bayesian deep reinforcement learning setup.
result Favorable experimental results on GridWorld and MiniGrid environments.
Paper proposes a framework for reliable off-policy evaluation in reinforcement learning.
problem Quantifying uncertainty in off-policy estimates for safe deployment of target policies.
method Distributionally robust optimization for creating confidence bounds.
result Non-asymptotic and asymptotic guarantees for robust cumulative reward estimates.
The paper tackles personalized policy learning from diverse data sources in a federated setting.
problem Learning personalized decision policies from observational bandit feedback across multiple heterogeneous data sources.
method Introduces a novel regret analysis for distinguishing global and local regret, and presents a federated policy learning algorithm using local policies trained with doubly robust offline policy evaluation strategies.
result Establishes finite-sample upper bounds on global and local regret, characterizing them by source heterogeneity and distribution shift.
We propose expected policy gradients (EPG), which unify stochastic policy gradients (SPG) and deterministic policy gradients (DPG) for reinforcement learning. Inspired by expected sarsa, EPG integrates (or sums) across actions when estimating the gradient, instead of relying only on the action in the sampled trajectory…
A new policy learning method allows policies to abstain when uncertain, improving safety and applicability.
problem Risk of making decisions without full confidence in uncertain predictions.
method Policy learning with abstention, identifying near-optimal policies and constructing an abstention rule.
result Improved safety and applicability in policy learning, with theoretical guarantees.
We study reinforcement learning of chatbots with recurrent neural network architectures when the rewards are noisy and expensive to obtain. For instance, a chatbot used in automated customer service support can be scored by quality assurance agents, but this process can be expensive, time consuming and noisy. Previous …
This paper analyzes Thompson Sampling in restless bandits with unknown parameters.
problem Analyzing performance of Thompson Sampling in restless bandits with unknown parameters.
method Proved a Bayesian regret bound of i l d e O ( T ) ilde{\mathcal{O}}(\sqrt{T}) i l d e O ( T ) for Thompson Sampling in episodic restless bandits with unknown parameters. result Proved a regret bound of i l d e O ( T ) ilde{\mathcal{O}}(\sqrt{T}) i l d e O ( T ) for Thompson Sampling in restless bandits with unknown parameters. Paper uses RL to optimize ICU load during COVID-19.
problem Optimizing ICU load during a pandemic.
method Combines epidemic model, Bayesian inference, and RL for adaptive intervention levels.
result RL policies reduce ICU burden compared to historical interventions.
CPME embeds counterfactual outcomes in RKHS for flexible policy evaluation.
problem Estimating counterfactual policy outcomes for decision-making.
method Counterfactual Policy Mean Embedding (CPME) framework in RKHS, plug-in and doubly robust estimators, kernel test statistic.
result Doubly robust estimator improves convergence rates and asymptotic normality.
GRAPE improves RL policy evaluation in noisy environments.
problem Noise in real-world RL environments makes policy evaluation algorithms inefficient or prone to errors.
method GRAPE combines gap-increasing value update operators and off-policy eligibility trace.
result GRAPE is more efficient and noise-tolerant than existing methods.
The paper develops a method to learn robust decision policies from observational data, reducing high-cost outcomes.
problem Learning safe decision policies from observational data with high-risk outcomes.
method Develops a method to learn policies that reduce high-cost outcomes, valid under finite samples and uneven feature overlap.
result Validates the method with real and synthetic data, providing statistical bounds on decision costs.
Paper explores models for summarizing AI agent policies.
problem Improving human understanding of AI agent behavior.
method Imitation learning-based approach to policy summarization.
result Matching summary extraction model to user model improves performance.
Decision-makers are faced with the challenge of estimating what is likely to happen when they take an action. For instance, if I choose not to treat this patient, are they likely to die? Practitioners commonly use supervised learning algorithms to fit predictive models that help decision-makers reason about likely futu…
A new multi-agent learning method improves performance in complex games.
problem Performance gap between MAPG and value-based multi-agent approaches.
method Introduces value function decomposition into multi-agent actor-critic framework for off-policy learning.
result DOP significantly outperforms state-of-the-art multi-agent reinforcement learning algorithms.
Paper tackles multi-view reinforcement learning with two methods.
problem Decision making with shared dynamics and different observation models.
method Observation augmentation and cross-view policy transfer.
result Reductions in sample complexities and computational time for multi-view environments.
Develops a method to estimate policy values robustly in the presence of confounding variables.
problem Infinite-horizon reinforcement learning with unobserved confounding variables makes policy evaluation unidentifiable.
method Robust approach estimating sharp bounds on policy value using optimization over state-occupancy ratios and sensitivity model.
result Proves convergence to sharp bounds as more confounded data is collected.
Simplifies BCQ to match and outperform state-of-the-art in offline RL benchmarks.
problem Sample efficiency in offline reinforcement learning.
method Introduces EMaQ, a novel backup operator for offline RL.
result EMaQ matches and outperforms prior state-of-the-art in offline RL benchmarks.
ZDPG learns model-free policies without critics, improving on PG.
problem Model-free policy learning in complex dynamic problems.
method Approximates policy-reward gradients via two-point stochastic evaluations of the Q-function.
result Restores true model-free policy learning without critics, with improved stability and efficiency.
A new framework for offline RL improves policy flexibility and regularity.
problem Lack of environmental interactions in offline RL leads to poor policy performance.
method Proposes a behavior-regularized implicit policy framework with modified policy-matching methods.
result The framework improves policy effectiveness and robustness beyond static datasets.
Automates infectious disease policy-making via inference in epidemiological models.
problem Improving policy-making for infectious diseases during pandemics.
method Performing inference in existing epidemiological models using a probabilistic programming language.
result Automated inference leads to better disease progression outcomes and policy prescriptions.