Zeroth-Order Non-Convex Learning via Hierarchical Dual Averaging

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

ICML 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Numerical experiments. For illustration purposes, Fig. 3 provides some numerical experiments on different no-regret policies discussed in the rest of our paper.
Researcher Affiliation Collaboration Amélie Héliou 1 Matthieu Martin 1 Panayotis Mertikopoulos 2 1 Thibaud Rahier 1 Criteo AI Lab 2Univ. Grenoble Alpes, CNRS, Inria, Grenoble INP, LIG, 38000 Grenoble, France.
Pseudocode No The paper defines algorithms like DAX and HDA through mathematical equations and descriptions of their components, but it does not present them in a formally structured pseudocode or algorithm block.
Open Source Code No The paper does not provide any explicit statement about making its source code available, nor does it provide a link to a code repository.
Open Datasets No The paper does not mention using a specific named public dataset for its numerical experiments. It describes the adversarial function as "analytic and randomly drawn" which suggests a synthetic setup rather than a public dataset.
Dataset Splits No The paper does not mention using specific training, validation, or test dataset splits. This information is typically provided when empirical evaluation is performed on well-defined datasets, which are not detailed here.
Hardware Specification No The paper does not specify any hardware details (e.g., GPU/CPU models, memory) used for running its experiments.
Software Dependencies No The paper does not provide specific software dependencies, such as library names with version numbers, that would be needed to replicate the experiments.
Experiment Setup No The paper mentions that "First the Hierarchical method is as outlined in Section 4 with parameters of the algorithm described below" and "We present the full details of our experiments in Appendix D." However, the provided text does not include Appendix D or the specific parameter values, making it impossible to determine concrete setup details from this excerpt.