---------------------------------------------------
== Call for Papers and Demos ==

ICML-2010 Workshop on BUDGETED LEARNING
Haifa, Israel
June 25, 2010

Submission Deadline: May 8, 2010
Notification of acceptance: May 23, 2010
---------------------------------------------------

Workshop home: http://www.dmargineantu.net/workshop/BudgetedLearning_ICML-2010

We would like to invite submissions to the ICML-2010 workshop on BUDGETED LEARNING. The workshop will mainly focus on algorithms and techniques that are required to learn on a fixed budget in the presence of non-uniform cost functions for the acquisition of feature values, labels, or entire instances, and for prediction errors. The organizers are seeking high quality contributions describing recent or ongoing work in the scope of Budgeted Learning. Both theoretical and applied work is solicited; demonstrations of experimental and/or deployed systems are especially encouraged.

Research questions and areas of interest include:
* Active learning with limited labeling resources
* Incorporating feature acquisition costs into learning algorithms
* Handling false alarm budgets in anomaly detection
* Utility/cost structures for acquiring individual features of instances
* Planning and search with limited resources
* Online learning where learner on a budget chooses to use and pay for the instances
* Budgeted learning of policies and sequential decision models (RL, KWIK learning, online learning)
* Learning with limited resources in adversarial settings and limited rationality
* Applications and their requirements for budgeted learning
* Evaluation and testing issues related to budgeted learning and decision systems
* Intelligent user interfaces that facilitate labeling and lower label/feature cost
* Labeled data as a commodity - Mechanical Turk and market design for labeling


Workshop Organizers
===================
Dragos Margineantu, Boeing Research & Technology (dragos.d.margineantu@boeing.com)
Russell Greiner, University of Alberta
Tomas Singliar, Boeing Research & Technology
Prem Melville, IBM Research (pmelvil@us.ibm.com)