Online Non-Convex Optimization with Imperfect Feedback

Authors: Amélie Héliou, Matthieu Martin, Panayotis Mertikopoulos, Thibaud Rahier

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

Reproducibility Variable Result LLM Response
Research Type Theoretical We consider the problem of online learning with non-convex losses. In this general context, we derive a series of tight regret minimization guarantees, both for the learner s static (external) regret, as well as the regret incurred against the best dynamic policy in hindsight. Subsequently, we apply this general template to the case where the learner only has access to the actual loss incurred at each stage of the process.
Researcher Affiliation Collaboration Amélie Héliou Criteo AI Lab a.heliou@criteo.com Matthieu Martin Criteo AI Lab mat.martin@criteo.com Panayotis Mertikopoulos Univ. Grenoble Alpes, CNRS, Inria, LIG & Criteo AI Lab panayotis.mertikopoulos@imag.fr Thibaud Rahier Criteo AI Lab t.rahier@criteo.com
Pseudocode Yes Algorithm 1: Dual averaging with imperfect feedback [Hedge variant: Q Λ] Algorithm 2: Bandit dual averaging [Hedge variant: Q Λ]
Open Source Code No The paper is theoretical and focuses on mathematical derivations and algorithm design. It does not contain any statements or links indicating the availability of source code for the described methods.
Open Datasets No The paper is theoretical and does not involve experimental evaluation using datasets. It provides illustrative examples but does not refer to any specific publicly available datasets for training.
Dataset Splits No The paper is theoretical and does not describe experimental validation or dataset splits (training, validation, or testing).
Hardware Specification No The paper is theoretical and does not describe any experiments that would require specific hardware. Therefore, no hardware specifications are mentioned.
Software Dependencies No The paper is theoretical and does not describe any experimental implementation details or specific software dependencies with version numbers.
Experiment Setup No The paper is theoretical and focuses on algorithm design and mathematical analysis. It does not describe an experimental setup with hyperparameters or system-level training settings.