Derivative-Free Optimization of High-Dimensional Non-Convex Functions by Sequential Random Embeddings

Authors: Hong Qian, Yi-Qi Hu, Yang Yu

IJCAI 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We apply SRE to several state-of-the-art derivative-free optimization methods, and conduct experiments on synthetic functions as well as non-convex classification tasks with up to 100,000 variables. Experiment results verify the effectiveness of SRE.
Researcher Affiliation Academia Hong Qian, Yi-Qi Hu, and Yang Yu National Key Laboratory for Novel Software Technology, Nanjing University, Nanjing 210023, China {qianh,huyq,yuy}@lamda.nju.edu.cn
Pseudocode Yes Algorithm 1 Sequential Random Embeddings (SRE)
Open Source Code No The paper mentions implementations of third-party methods are 'by their authors' but does not provide any link or statement for the availability of their own SRE code.
Open Datasets Yes We employ four binary class UCI datasets [Blake et al., 1998], Gisette, Arcene, Dexter and Dorothea.
Dataset Splits No The paper uses terms like 'training instance' and mentions testing, but it does not specify explicit percentages or counts for training, validation, and test splits, nor does it provide a splitting methodology or reference predefined splits for reproducibility.
Hardware Specification No No specific hardware details (like GPU/CPU models, memory, or specific computing environments) used for running experiments are mentioned in the paper.
Software Dependencies No The paper mentions software like IMGPO, CMAES, and RACOS but does not provide specific version numbers for these or any other software dependencies.
Experiment Setup Yes We set D = 10000, set the total number of function evaluations n = 10000 and the subspace size d = 10, and choose the number of sequential random embeddings m = {1, 2, 5, 8, 10, 20}. ... We set d = 20, the number of function evaluations n = 3D, and X = [ 10, 10]D, Y = [ 10, 10]d, 2 [ 10, 10] for all applied algorithms except for CCCP. ... For algorithms with SRE, we set the number of sequential random embeddings m = 5.