Explicit Gradient Learning for Black-Box Optimization

Authors: Elad Sarafian, Mor Sinay, Yoram Louzoun, Noa Agmon, Sarit Kraus

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

Reproducibility Variable Result LLM Response
Research Type Experimental We evaluate EGL and achieve state-of-the-art results in two challenging problems: (1) the COCO test suite against an assortment of standard BBO methods; and (2) in a high-dimensional non-convex image generation task.
Researcher Affiliation Academia 1Department of Computer Science, Bar Ilan University, Israel. Correspondence to: Elad Sarafian, Mor Sinay <elad.sarafian@gmail.com, mor.sinay@gmail.com>
Pseudocode Yes Algorithm 1 Indirect Gradient Learning (...) Algorithm 2 Convergent EGL
Open Source Code Yes The code is available at http://github.com/Mor Sinay/BBO.
Open Datasets Yes We tested EGL on the COCO test suite (...) We trained a generator & discriminator for the Celeb A dataset (Liu et al., 2015)
Dataset Splits No The paper does not provide specific train/validation/test dataset splits, percentages, or references to predefined splits needed to reproduce data partitioning.
Hardware Specification No The paper does not explicitly describe any specific hardware components (e.g., GPU/CPU models, memory details) used for running the experiments.
Software Dependencies No The paper mentions using 'Scipy and Cma Python packages' but does not provide specific version numbers for these or other software dependencies.
Experiment Setup Yes For additional information, including a hyperparameters list, refer to Appendix Sec. F.