Patent · US Active

Accelerated discrete distribution clustering under wasserstein distance

US10013477B2 · kind B2 · utility

4Cited by

6References

5Claims

0Family size

Assignee

The Penn State Research Foundation · US

Inventors

Jianbo Ye · Beijing, CN
Jia Li · State College, US
James Z. Wang · State College, US

Key dates

Filing date	Sep 30, 2016
Grant date	Jul 3, 2018
Priority date	—
Expiry date	Sep 30, 2036

Classification

Technology area (CPC G)Physics
CPC primaryG06N20/10
WIPO fieldComputer technology
WIPO sectorElectrical engineering

Abstract

Computationally efficient accelerated D2-clustering algorithms are disclosed for clustering discrete distributions under the Wasserstein distance with improved scalability. Three first-order methods include subgradient descent method with re-parametrization, alternating direction method of multipliers (ADMM), and a modified version of Bregman ADMM. The effects of the hyper-parameters on robustness, convergence, and speed of optimization are thoroughly examined. A parallel algorithm for the modified Bregman ADMM method is tested in a multi-core environment with adequate scaling efficiency subject to hundreds of CPUs, demonstrating the effectiveness of AD2-clustering.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.