Reducing Adversarially Robust Learning to Non-Robust PAC Learning

Authors: Omar Montasser, Steve Hanneke, Nati Srebro

NeurIPS 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Theoretical We study the problem of reducing adversarially robust learning to standard PAC learning, i.e. the complexity of learning adversarially robust predictors using access to only a black-box non-robust learner. We give a reduction that can robustly learn any hypothesis class C using any non-robust learner A for C. The number of calls to A depends logarithmically on the number of allowed adversarial perturbations per example, and we give a lower bound showing this is unavoidable.
Researcher Affiliation Academia Omar Montasser omar@ttic.edu Steve Hanneke steve.hanneke@gmail.com Nathan Srebro nati@ttic.edu Toyota Technological Institute at Chicago
Pseudocode Yes Algorithm 1: Robustify The Non-Robust
Open Source Code No The paper does not provide any links to open-source code or state that code will be released.
Open Datasets No The paper is theoretical and does not involve empirical training on specific datasets.
Dataset Splits No The paper is theoretical and does not involve empirical validation on specific datasets, hence no training/validation/test splits are mentioned.
Hardware Specification No The paper is theoretical and does not mention any hardware specifications used for experiments.
Software Dependencies No The paper is theoretical and does not mention specific software dependencies with version numbers for experiments.
Experiment Setup No The paper is theoretical and does not describe an experimental setup with hyperparameters or training settings.