On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

168,657 papers · 148 categories

Trend · papers per month

count % of corpus

9 results for “uniform-PAC”

Uniform-PAC Guarantees for Model-Based RL with Bounded Eluder Dimensioncs.LG

New algorithms achieve uniform-PAC guarantees for RL with bounded eluder dimension.

problem Achieving strong performance guarantees in reinforcement learning.

method Proposes algorithms for nonlinear bandits and model-based episodic RL with a bounded eluder dimension.

result Achieves uniform-PAC sample complexity that matches state-of-the-art regret bounds or sample complexity guarantees.

reinforcement learning uniform-PAC guarantees model-based RL bounded eluder dimension nonlinear bandits

2023-05-15abs ↗pdf ↗

Uniform-PAC Bounds for Reinforcement Learning with Linear Function Approximationcs.LG

New algorithm FLUTE achieves uniform-PAC convergence in RL with linear approx.

problem RL with linear function approximation lacks uniform-PAC guarantees.

method FLUTE algorithm with minimax value function estimator and multi-level partition scheme.

result Uniform-PAC convergence to optimal policy with high probability.

reinforcement learning linear function approximation uniform-PAC minimax value function multi-level partition

2021-06-22abs ↗pdf ↗

Unifying PAC and Regret: Uniform PAC Bounds for Episodic Reinforcement Learningcs.LG

Statistical performance bounds for reinforcement learning (RL) algorithms can be critical for high-stakes applications like healthcare. This paper introduces a new framework for theoretically measuring the performance of such algorithms called Uniform-PAC, which is a strengthening of the classical Probably Approximatel…

2017-03-22abs ↗pdf ↗

A unified recipe for deriving (time-uniform) PAC-Bayes boundsstat.ML

Unified framework for anytime-valid PAC-Bayes bounds.

problem Deriving time-uniform PAC-Bayes bounds for stochastic processes.

method Combines four tools: nonnegative supermartingales, method of mixtures, Donsker-Varadhan formula, and Ville's inequality.

result Unified PAC-Bayes theorem for a wide class of discrete stochastic processes.

PAC-Bayes time-uniform bounds stochastic processes supermartingales mixtures

2023-02-07abs ↗pdf ↗

Testing for Overfittingstat.ML

Tests for overfitting in machine learning models.

problem Overfitting in high complexity models.

method Hypothesis test using concentration bounds.

result Valid test for identifying overfitting.

overfitting machine learning empirical risk concentration bounds model generalization

2023-05-09abs ↗pdf ↗

A Unified Confidence Sequence for Generalized Linear Models, with Applications to Banditsstat.ML

Unified CS for GLMs improves bandit regret bounds.

problem Improving regret bounds for GLMs in bandit settings.

method Unified likelihood ratio-based CS with PAC-Bayesian bound.

result Unified CS attains poly(S)-free regret for Bernoulli.

Generalized Linear Models Confidence Sequences Bandits Optimistic Algorithms PAC-Bayesian

2024-07-19abs ↗pdf ↗

PAC Guarantees for Reinforcement Learning: Sample Complexity, Coverage, and Structurecs.LG

Survey of reinforcement learning guarantees with data constraints.

problem Guaranteeing near-optimal policies with limited data in reinforcement learning.

method Coverage-Structure-Objective (CSO) framework to decompose sample complexity results.

result Progress on PAC guarantees for reinforcement learning, covering various models and settings.

PAC guarantees reinforcement learning sample complexity coverage structure

2026-03-01abs ↗pdf ↗

Uniform Last-Iterate Guarantee for Bandits and Reinforcement Learningcs.LG

This paper introduces a new metric, ULI, for RL that ensures both cumulative and instantaneous performance.

problem High-stakes applications require RL algorithms to avoid playing bad policies.

method Introduces uniform last-iterate (ULI) guarantee, a stronger metric capturing both cumulative and instantaneous performance.

result ULI directly implies near-optimal cumulative performance across various metrics, but not the other way around.

reinforcement learning uniform last-iterate bandits optimism cumulative performance

2024-02-20abs ↗pdf ↗

Frozen Policy Iteration: Computationally Efficient RL under Linear

Q^π

Realizability for Deterministic Dynamicscs.LG

Efficient RL algorithm for MDPs with linear $Q^π$ realizability, achieving optimal regret bound.

problem Efficient reinforcement learning under linear

Q^π

realizability assumption for MDPs with stochastic dynamics.

method Frozen Policy Iteration algorithm that uses high-confidence data and freezes policy for well-explored states.

result Achieves optimal regret bound of

\widetilde{O}(\sqrt{d^2H^6T})

for linear (contextual) bandits.

reinforcement learning Markov Decision Processes linear realizability online learning policy iteration

2026-02-28abs ↗pdf ↗