Can Stochastic Zeroth-Order Frank-Wolfe Method Converge Faster for Non-Convex Problems?

Authors: Hongchang Gao, Heng Huang

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

Reproducibility Variable Result LLM Response
Research Type Experimental At last, the empirical studies on benchmark datasets validate our theoretical results. 4. Experiments
Researcher Affiliation Collaboration 1Department of Electrical and Computer Engineering, University of Pittsburgh, Pittsburgh, USA 2Department of Computer and Information Sciences, Temple University, Philadelphia, USA 3JD Finance America Corporation.
Pseudocode Yes Algorithm 1 Faster Zeroth-Order Frank-Wolfe Method (FZFW) ... Algorithm 2 Faster Zeroth-Order Conditional Gradient Method (FZCGS) ... Algorithm 3 u+ = condg(g, u, γ, ) (Qu et al., 2017) ... Algorithm 4 Faster First-Order Conditional Gradient Sliding Method (FCGS)
Open Source Code No The paper includes a footnote linking to a GitHub repository (https://github.com/IBM/ ZOSVRG-Black Box-Adv) for a pre-trained DNN model used in one experiment, but it does not state that the source code for the proposed methods is available.
Open Datasets Yes The dataset used in this experiment is Movie Lens100k... Following (Liu et al., 2018; Ji et al., 2019), we use the same pretrained DNN2 for MNIST dataset as the black-box model.
Dataset Splits No The paper describes synthetic data generation and mentions MovieLens100k and MNIST datasets, but it does not provide specific details on train/validation/test splits (e.g., percentages or sample counts) for any of these datasets.
Hardware Specification No The paper does not provide specific hardware details such as GPU models, CPU types, or memory specifications used for running the experiments.
Software Dependencies No The paper does not provide specific software dependencies with version numbers (e.g., Python 3.x, PyTorch 1.x) that would be needed to reproduce the experiments.
Experiment Setup Yes The hyperparameter s is set to 0.1.