[November 24th, 2009]  Workshop program is posted online.

[November 24th, 2009] Confirmed Invited Speaker:  Prof. Sanjay Ranka (University of Florida).  Talk title:  Novel Mixture Models to Learn Complex Patterns in High-Dimensional Data.
 

[November 10th, 2009] Confirmed Invited Speaker: Prof. Chris Ding (University of Texas at Arlington). Talk title: Tensor Decompositions: New Clustering Capability and Error Bounds.
 

[October 20, 2009] The list of accepted papers is listed below:

• Chengying Miao. An Effective Network Partitioning Algorithm Based on Two-Point Diffusing Strategy
• Dan Zhang and Luo Si. Multiple Instance Transfer Learning.
• Pritam Chanda, Aidong Zhang and Murali Ramanathan. Mining of Attribute Interactions Using Information Theoretic Metrics.
• Shuo Chen, Bin Liu, Mingjie Qian and Changshui Zhang. Kernel K-means Based Framework for Aggregate Outputs Classification.
• Ke Tang and Rui Wang. Feature Selection for Maximizing the Area Under the ROC Curve.
• Jana Nononicova, Petr Somol and Pavel Pudi. A New Stability Measure for Feature Selection Algorithms.
• Hongliang Fei, Brian Quanz and Jun Huan. GLSVM: Integrating Structured Feature Selection and Large Margin Classification.
• Mingjie Qian, Feiping Nie and Changshui Zhang. Probabilistic Labeled Semi-supervised SVM.
• Kunal Punera and Suju Rajan. Improving Multilabel Classification in Hierarchical Taxonomies.
• Gregory Moore, Charles Bergeron and Kristin Bennett. Nonconvex Bilevel Programming for Hyperparameter Selection.

[June 1st, 2009] PC members added

[May 16th, 2009]  Webpage Kickoff

 

Classical optimization techniques have found widespread use in solving  traditional data mining problems, among which convex optimization has occupied
the center-stage because of its elegant property of global optimum. Many problems can be casted into the convex optimization framework, such as Support
Vector Machines, graph-based manifold learning, and clustering, which can usually be solved by convex Quadratic Programming, Semi-Definite Programming or
Eigenvalue Decomposition.

As time goes by, new problems emerge constantly in data mining community, such  as Time-Evolving Data Mining, On-Line Data Mining, Relational Data Mining and Transferred Data Mining. While at the same time fundamental problems such as classification and clustering continue to be better understand. Some of these
recently emerged problems are more complex than traditional ones and are usually  formulated as nonconvex problems. Therefore some general optimization methods, such as gradient descents, coordinate descents, convex relaxation, have come  back to the stage and become more and more popular in recent years.

This workshop will present recent advances in optimization techniques for, especially new emerging, data mining problems, as well as the real-life
applications among this community. One main goal of the workshop is to bring  together leading researchers who work on state-of-the-art algorithms on
optimization based methods for modern data analysis, and also the practitioners  who seek for novel applications. In summary, this workshop will strive to emphasize the following aspects:

• Presenting recent advances in algorithms and methods using optimization techniques
• Addressing the fundamental challenges in data mining using optimization techniques
• Identifying killer applications and key industry drivers (where theories and applications meet)
• Fostering interactions among researchers (from different backgrounds) sharing the same interest to promote cross-fertilization of ideas.
• Exploring benchmark data for better evaluation of the techniques