Online Estimation via Offline Estimation: An Information-Theoretic Framework

Authors: Dylan J Foster, Yanjun Han, Jian Qian, Alexander Rakhlin

NeurIPS 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Theoretical Online Estimation via Offline Estimation: An Information-Theoretic Framework. Our main results settle the statistical and computational complexity of online estimation in this framework. Finally, we apply our results to give offline oracle-efficient algorithms for interactive decision making.
Researcher Affiliation Collaboration Dylan J. Foster dylanfoster@microsoft.com Yanjun Han yanjunhan@nyu.edu Jian Qian jianqian@mit.edu Alexander Rakhlin rakhlin@mit.edu
Pseudocode Yes Algorithm 1 Version Space Averaging; Algorithm 2 Reduction from OEOE to Online Learning with Delayed Feedback; Algorithm 3 Estimation to Decisions Meta-Algorithm with Offline Oracles (E2D.Off); Algorithm 4 Reduction to delayed online learning for binary loss; Algorithm 5 Reduction from delayed online learning to non-delayed online learning; Algorithm 6 Reduction from CDEw RO to CDE; Algorithm 7 Reduction from CDEw RP to CDEw RO; Algorithm 8 Reduction from CDEw DRP to CDEw RP; Algorithm 9 Reduction from OEOE to CDEw DRP
Open Source Code No The paper does not contain any statement about making its code open-source or providing a link to a code repository.
Open Datasets No The paper is theoretical and discusses abstract 'instances' and 'classes' (e.g., '(X, Y, Z, K, F)') without referring to specific, named, or publicly accessible datasets.
Dataset Splits No The paper is theoretical and does not report on experiments with dataset splits.
Hardware Specification No The paper is theoretical and does not report on experiments that would require hardware specifications.
Software Dependencies No The paper is theoretical and does not report on experiments that would require specific software dependencies or versions.
Experiment Setup No The paper is theoretical and does not report on experimental setups or hyperparameters.