About Me
I received my Ph.D. from the School of Industrial Engineering at Purdue University in 2021, under the supervision of Prof. Gesualdo Scutari.
Previously, I got my B.S. from the Department of Mathematics at Nanjing University in 2016.
My research interest lies in designing efficient and robust decentralized/distributed optimization algorithms and their application to machine learning.
Selected Publications
-
Acceleration in Distributed Optimization under Similarity
Ye Tian, Gesualdo Scutari, Tianyu Cao, and Alexander Gasnikov,
International Conference on Artificial Intelligence and Statistics (AISTATS) 2022. [PDF]
This paper presents an accelerated decentralized optimization algorithm whose communication complexity is optimal up to log factors, in the presence of similarity among local datasets.
-
ASY-SONATA: Achieving Geometric Convergence for Distributed Asynchronous Optimization
Ye Tian, Ying Sun, and Gesualdo Scutari,
Allerton Conference on Communication, Control, and Computing (Allerton) 2018.
[PDF] [code]
1) This paper presents a decentralized asynchronous (first-order) optimization algorithm achieving geometric convergence rate for minimizing strongly convex objectives, and sublinear rate for general nonconvex problems.
2) We adopt the partially asynchronous algorithmic model, which is mathematically general and does not enforce any specific computation/communication protocol.
-
On coding capacity of delay-constrained network information flow: An algebraic approach
Minghua Chen, Ye Tian, and Chih-Chun Wang,
IEEE International Symposium on Information Theory (ISIT) 2016.
[PDF]
This paper characterizes the communication capacity of Linear Network Coding in multicast/broadcast scenario with a hard delay constraint.