Special Session Call for Papers
Learning from Heterogeneous Data
13-15 Dec. 2009,
AIMS AND SCOPE
Scientists are increasingly facing with emerging challenges arose from the dramatic growth in the availability of data from a variety sources. For example, image classification tasks can use data collected from different sensors; bioinformatics tasks can exploit protein sequences, gene expression profiles and ontologies. These heterogeneous data sets allow a researcher to leverage different characteristics of a sample that one homogenous data set cannot represent. Empirical experiments have shown that using heterogeneous features allows for significant gains in performance.
While these heterogeneous data sets can be found in a variety of machine learning applications, including biomedicine, web mining, target detection, and business, to name a few, standard machine learning algorithms may be limited by the general underlying assumptions that the training data available are drawn from a single source and that each sample is represented by a single vector of variables. Therefore, more sophisticated learning and data fusion methods are necessary to make the best use possible of heterogeneous data sets. Many novel approaches have been developed in the last ten years, but many of these are specific to the application; generative models may be useful in a wide variety of fields. The goal of this special session is to share the approaches used by researchers studying various applications and to encourage state-of-the-art research in learning from heterogeneous data sources. The topics range from case studies of particular problems with heterogeneous data to novel learning approaches such as kernel methods, ensemble approaches and feature selection/extraction methods.
Paper Submission Deadline: August 1, 2009
Notification of acceptance: September 7, 2009
Camera-ready papers & Pre-registration: October 1, 2009
The ICMLA Conference: December 13-15, 2009
The authors would submit papers through the main conference submission website. Papers must correspond to the requirements detailed in the instructions to authors. Accepted papers must be presented by one of the authors to be published in the conference proceeding. If you have any questions, do not hesitate to direct your questions to Dr. Chen at email: firstname.lastname@example.org.
All paper submissions will be handled electronically. Detailed instructions for submitting the papers are provided on the conference home page at
Special Session Co-chairs:
· Xue-wen Chen, The University of Kansas
Program Committee Members:
· Xue-wen Chen,
· Christophe Giraud-Carrier,
· Haiyan Huang,
· Zicheng Liu, Microsoft Research
· Volker Willert , Technical
· Dit-Yan Yeung,