Replica Exchange for Non-Convex Optimization

Authors: Jing Dong, Xin T. Tong

JMLR 2021 | Venue PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We further verify our theoretical results through some numerical experiments, and observe superior performance of the proposed algorithm over running GD or LD alone. In this section, we provide some numerical experiments to illustrate the performance of (n)GDx LD and (n)SGDx SGLD.
Researcher Affiliation Academia Jing Dong EMAIL Graduate School of Business Columbia University 3022 Broadway New York, NY 10027, USA Xin T. Tong EMAIL Department of Mathematics National University of Singapore Block S17, 10 Lower Kent Ridge Road Singapore 119077
Pseudocode Yes Algorithm 1: GDx LD and n GDx LD: offline optimization... Algorithm 2: SGDx SGLD and n SGDx SGLD: online optimization
Open Source Code No The paper does not provide an explicit statement or a link to an open-source code repository for the described methodology.
Open Datasets No The paper describes the generation of data for its numerical experiments, such as 'Gaussian-mixture density' and 'mixture of Gaussian distribution', and uses mathematical test functions like 'Rastrigin function' and 'Griewank function'. It does not explicitly provide access information (links, DOIs, or citations) for any pre-existing public datasets.
Dataset Splits No The paper uses generated data and mathematical test functions and does not discuss specific training/test/validation dataset splits.
Hardware Specification No The paper does not provide specific hardware details (e.g., GPU/CPU models, memory) used for running its experiments.
Software Dependencies No The paper does not provide specific software or library version numbers used for implementing the algorithms or conducting experiments.
Experiment Setup Yes We implement GDx LD for the objective function plotted in Figure 2 with h = 0.1, γ = 1, X0 = (0, 0), and Y0 = (1, 1). ... We set h = 0.1, γ = 1, Θ = 103, t0 = 0.05, Mv = 5, X0 = (0, 0), and Y0 = (1, 1).