Abstract
In this paper, we consider the use of structure learning methods for probabilistic graphical models to identify statistical dependencies in high-dimensional physical processes. Such processes are often synthetically characterized using PDEs (partial differential equations) and are observed in a variety of natural phenomena. In this paper, we present ACLIME-ADMM, an efficient two-step algorithm for adaptive structure learning, which decides a suitable edge specific threshold in a data-driven statistically rigorous manner. Both steps of our algorithm use (inexact) ADMM to solve suitable linear programs, and all iterations can be done in closed form in an efficient block parallel manner. We compare ACLIME-ADMM with baselines on both synthetic data simulated by PDEs that model advection-diffusion processes, and real data of daily global geopotential heights to study information flow in the atmosphere. ACLIME-ADMM is shown to be efficient, stable, and competitive, usually better than the baselines especially on difficult problems. On real data, ACLIME-ADMM recovers the underlying structure of global atmospheric circulation, including switches in wind directions at the equator and tropics entirely from the data.
Original language | English (US) |
---|---|
Title of host publication | Proceedings - 17th IEEE International Conference on Data Mining, ICDM 2017 |
Editors | George Karypis, Srinivas Alu, Vijay Raghavan, Xindong Wu, Lucio Miele |
Publisher | Institute of Electrical and Electronics Engineers Inc. |
Pages | 883-888 |
Number of pages | 6 |
ISBN (Electronic) | 9781538638347 |
DOIs | |
State | Published - Dec 15 2017 |
Event | 17th IEEE International Conference on Data Mining, ICDM 2017 - New Orleans, United States Duration: Nov 18 2017 → Nov 21 2017 |
Publication series
Name | Proceedings - IEEE International Conference on Data Mining, ICDM |
---|---|
Volume | 2017-November |
ISSN (Print) | 1550-4786 |
Other
Other | 17th IEEE International Conference on Data Mining, ICDM 2017 |
---|---|
Country/Territory | United States |
City | New Orleans |
Period | 11/18/17 → 11/21/17 |
Bibliographical note
Publisher Copyright:© 2017 IEEE.
Keywords
- ACLIME-ADMM
- Geoscience
- High-dimensional physical process
- PC stable
- Structure learning