Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].
Prediction against a limited adversary
Authors: Erhan Bayraktar, Ibrahim Ekren, Xin Zhang
JMLR 2021 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Theoretical | We study the problem of prediction with expert advice with adversarial corruption where the adversary can at most corrupt one expert. Using tools from viscosity theory, we characterize the long-time behavior of the value function of the game between the forecaster and the adversary. We provide lower and upper bounds for the growth rate of regret without relying on a comparison result. We show that depending on the description of regret, the limiting behavior of the game can significantly differ. [...] The rest of the paper is organized as follows. In Section 2, we formulate the problem of prediction against a limited adversary and state the relevant assumptions. In Section 3, we heuristically derive the limiting equation and in Section 4 state our main results. The Section 5 contains special cases where we can explicitly solve the limiting equation. |
| Researcher Affiliation | Academia | Erhan Bayraktar EMAIL Department of Mathematics University of Michigan Ann Arbor, MI 48109-1043, USA Ibrahim Ekren EMAIL Department of Mathematics Florida State Univeristy Tallahassee, FL 32306-4510, USA Xin Zhang EMAIL Department of Mathematics University of Michigan Ann Arbor, MI 48109-1043, USA |
| Pseudocode | No | The paper discusses 'algorithms for the forecaster and the adversary' and defines strategies mathematically (e.g., 'α m(x) = argmax α AB ...', 'φ m = { j U(tm 1, Xm 1)}N j=1'). However, these are presented as mathematical expressions or descriptions of optimal responses within the theoretical framework, not as structured, labeled pseudocode or algorithm blocks. There are no sections explicitly titled 'Algorithm' or 'Pseudocode'. |
| Open Source Code | No | The paper states 'License: CC-BY 4.0, see https://creativecommons.org/licenses/by/4.0/. Attribution requirements are provided at http://jmlr.org/papers/v22/20-1234.html.' This refers to the license for the paper itself, not for any accompanying source code. There is no explicit statement about releasing source code for the methodology described in the paper, nor any links to code repositories. |
| Open Datasets | No | This is a theoretical paper focusing on mathematical modeling and analysis of a prediction problem with an adversary. It does not conduct experiments with empirical data, and therefore, no datasets are used or made available. |
| Dataset Splits | No | This is a theoretical paper that does not involve empirical experiments with datasets. Therefore, there is no mention of dataset splits for training, validation, or testing. |
| Hardware Specification | No | This is a theoretical paper focused on mathematical analysis and proofs. It does not describe any computational experiments or their execution environment, and thus, no hardware specifications are provided. |
| Software Dependencies | No | This paper is theoretical in nature, presenting mathematical models, theorems, and solutions to partial differential equations. It does not describe any computational implementation or specific software requirements, nor does it list any software versions. |
| Experiment Setup | No | This paper is entirely theoretical, focusing on mathematical modeling and analysis. It does not present any experimental results from computational models, and therefore, there are no details regarding experimental setup, hyperparameters, or training configurations. |