posted by organizer: sanmik || 1730 views || tracked by 7 users: [display]

DMPI@ICML 2014 : ICML 2014 Workshop on Divergence Methods for Probabilistic Inference

FacebookTwitterLinkedInGoogle

Link: http://ml-thu.net/~dmpi-icml2014-workshop/home
 
When Jun 25, 2014 - Jun 26, 2014
Where Beijing, China
Submission Deadline Apr 13, 2014
Notification Due May 1, 2014
Final Version Due Jun 20, 2014
Categories    machine learning   statistics
 

Call For Papers

ICML 2014 Workshop on Divergence Methods for Probabilistic Inference
Submission Deadline: April 13, 2014
Beijing, June 25/26, 2014
http://ml-thu.net/~dmpi-icml2014-workshop/home

Call for Participation
-----------------------
Researchers in various sub-fields of machine learning and statistics such as natural language processing, collaborative filtering and neuroinformatics regularly employ probabilistic methods for analyzing data. Probabilistic inference is the process of updating a-priori uncertainty with new information such as new data samples and constraints, and is a cornerstone of modern machine learning and applied statistics. The workshop aims to explore divergence methods for probabilistic inference, considering issues such as the choice of divergence, the choice of constraints, efficient and scalable inference, and applications to big data. The workshop will provide a venue for researchers in disparate fields to interact, map out the state of the art in the field, and encourage discussion to stimulate new theoretical and practical developments.

Topics of interest include:
* Analysis: Analysis of the properties of different divergence functionals. Properties of the divergence based posterior estimates such as consistency, sample complexity and risk.
* Novel constraints: Development of novel approaches for incorporating constraints such as online constraint generation. Complex structured regularizers and graph regularizers such as wordnet for semantic relationships between name entities.
* Scalable inference for big data and complex data: Novel scalable inference methods for modern big data applications including parallel, distributed and streaming architectures for large scale inference. Novel models and inference methods for multi-relational and multi-modal data with complex interactions.
* Applications: Applications to various problem domains such as constrained clustering, natural language processing and scientific applications.

Submissions
------------
We invite submissions for original papers that introduce new research developments, directions, frameworks, results, etc. in these and related areas. We also invite submissions of recently published high-quality work. Though, we note that preference will be given to original contributions. Potential participants may submit full papers (up to 8 pages in length in ICML format) or short papers (extended abstracts, 2-4 pages in length) by April 11, 2014 sent electronically via the online submission site: https://cmt.research.microsoft.com/DMPI2014/.

The workshop is non-archival and the authors retain the copyrights to the entries and the rights to resubmit and publish at other venues.

Organizing Committee
---------------------
Sanmi Koyejo (University of Texas at Austin)
Jun Zhu (Tsinghua university)
Mark Reid (Australian National University)
Eric Xing (Carnegie Mellon University)

Related Resources

ICML 2017   34th International Conference on Machine Learning
DSAA 2017   The 4th IEEE International Conference on Data Science and Advanced Analytics 2017
PLP 2017   Probabilistic logic programming 2017
ICONIP 2017   International Conference on Neural Information Processing
LearnAut 2017   Learning and Automata - LICS 2017 Workshop
ACML 2017   The 9th Asian Conference on Machine Learning
ASPMMI 2017   IEEE Access: Special Issue on Advanced Signal Processing Methods in Medical Imaging
ICPR 2018   24th International Conference on Pattern Recognition
AutoML 2017   Automatic Machine Learning Workshop (ICML 2017)
MEMOCODE 2017   15th ACM/IEEE International Conference on Formal Methods and Models for System Design