26 Papers
130 Citations
Bin Shi is an academic researcher from University of California, Berkeley. The author has contributed to research in topics: Computer science & Convex function. The author has an hindex of 6, co-authored 18 publications. Previous affiliations of Bin Shi include Chinese Academy of Sciences & University of California.
Chat about Author
Papers
Understanding the acceleration phenomenon via high-resolution differential equations
TL;DR: An alternative limiting process that yields high-resolution ODEs permit a general Lyapunov function framework for the analysis of convergence in both continuous and discrete time and are more accurate surrogates for the underlying algorithms.
•Posted Content
Understanding the Acceleration Phenomenon via High-Resolution Differential Equations
TL;DR: In this paper, an alternative limiting process that yields high-resolution ODEs was proposed, which can be used to distinguish between Nesterov's accelerated gradient method for strongly convex functions (NAG-SC) and Polyak's heavy-ball method.
98
•Posted Content
On Learning Rates and Schrödinger Operators
TL;DR: This paper presents a general theoretical analysis of the effect of the learning rate in stochastic gradient descent (SGD), and provides a mathematical interpretation of the benefits of using learning rate decay for nonconvex optimization.
44
Risk-Resistant Resource Allocation for eMBB and URLLC Coexistence Under M/G/1 Queueing Model
Bin Shi,F. Zheng,Changyang She,Jingjing Luo,Alister G. Burr +4 more
TL;DR: This paper proposes a risk-resistant scheme for resource allocation to both eMBB and URLLC traffic that aims to maximize the e MBB data rate, while considering the URllC delay threshold violation to maintainURLLC delay requirements.
27
•Proceedings Article
Acceleration via Symplectic Discretization of High-Resolution Differential Equations
Bin Shi,Simon S. Du,Weijie J. Su,Michael I. Jordan +3 more
- 01 Feb 2019
TL;DR: It is shown that the optimization algorithm generated by applying the symplectic scheme to a high-resolution ODE proposed by Shi et al.