Acceleration and Averaging in Stochastic Descent Dynamics

Authors: Walid Krichene, Peter L. Bartlett

NeurIPS 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Theoretical We formulate and study a general family of (continuous-time) stochastic dynamics for accelerated first-order minimization of smooth convex functions. Building on an averaging formulation of accelerated mirror descent, we propose a stochastic variant in which the gradient is contaminated by noise, and study the resulting stochastic differential equation. We prove a bound on the rate of change of an energy function associated with the problem, then use it to derive estimates of convergence rates of the function values (almost surely and in expectation), both for persistent and asymptotically vanishing noise.
Researcher Affiliation Collaboration Walid Krichene Google, Inc. walidk@google.com Peter Bartlett U.C. Berkeley bartlett@cs.berkeley.edu
Pseudocode No The paper is highly theoretical and presents mathematical formulations and proofs, but it does not include pseudocode or algorithm blocks.
Open Source Code No The paper does not mention releasing any source code or provide links to a code repository.
Open Datasets No The paper is theoretical and does not involve empirical training on datasets. Therefore, no information about publicly available training data is provided.
Dataset Splits No The paper is theoretical and does not involve empirical experiments with data splits for validation.
Hardware Specification No The paper is theoretical and does not describe empirical experiments that would require hardware specifications.
Software Dependencies No The paper is theoretical and does not mention specific software dependencies with version numbers required for reproducibility of empirical experiments.
Experiment Setup No The paper is theoretical and does not describe an experimental setup with hyperparameters or training configurations.