A Dynamical Systems Perspective on Nesterov Acceleration

Authors: Michael Muehlebach, Michael Jordan

ICML 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Sec. 5 presents a simulation example that illustrates the properties of the continuous-time dynamics and the discretization.
Researcher Affiliation Academia 1Electrical Engineering and Computer Science Department, UC Berkeley, Berkeley, California, USA.
Pseudocode No The paper provides mathematical equations for discretization (e.g., equations 3, 4, 10, 11, 12, 13) but does not include a formal pseudocode block or algorithm environment.
Open Source Code No The paper does not contain any statements about releasing open-source code for the described methodology, nor does it provide any links to a code repository.
Open Datasets No The paper's simulation example uses a custom function 'f' defined by its gradient, rather than a publicly available or open dataset. 'We choose a function f with the following gradient'.
Dataset Splits No The paper describes a simulation example using a custom function, and therefore, does not provide dataset split information (e.g., percentages, sample counts, or citations to predefined splits) typically associated with machine learning datasets.
Hardware Specification No The paper does not provide any specific hardware details such as GPU models, CPU specifications, or cloud computing instances used for running the simulations.
Software Dependencies No The paper does not list any specific software dependencies or their version numbers (e.g., Python version, library names, or solver versions) used for the simulations.
Experiment Setup Yes We choose a function f with the following gradient... where the condition number κ is set to 5. The integration algorithm given by (3) and (4) is applied to the initial conditions (q0, 0), where q0 is varied from 2 to 5 in steps of 0.2. The step size Ts is successively increased from 0.1 to 1.2.