Special Session

on

Statistical Data Mining and Machine Learning in Cancer Epidemiology and Cancer Bioinformatics

Of

The Seventh International Conference on Machine Learning and Applications (ICMLA’ 08)

http://www.icmla-conference.org/icmla08/

December 11-13, 2008 – San Diego, California, USA

Motivation

High throughput technologies such as microarray have produced huge amount of data in public domain. Many survey and clinical outcome data such as SEER data are also available.  A long list of links to large health-related data sets can be found at the website http://www.ehdp.com/vitalnet/datasets.htm. All of these databases have different temporal and spatial assumptions (for example, different frequencies of collection, different spatial resolution (by state, by county, by zip-code, by square kilometer), etc. How to mine these data together and extract useful information is really a challenging task. Although we have seen many applications of data mining and machine learning techniques in microarray and other high throughput data, there are much less applications in SNP array and cancer epidemiology.  It is the organizer’s belief that new computational methods are needed to deal with large, complex data sets arising in cancer epidemiology and cancer bioinformatics. We see a pressing need for and benefits in the interdisciplinary exchange and discussion of ideas. We anticipate that this workshop will shed light on research directions and provide the stimulus for creative breakthroughs.

Themes

This special session will bring together researchers from different disciplines and encourage collaborative research on cancer related data mining. The objectives of this workshop are intended to addressing two challenging issues.  One is how to identify and evaluate biomarkers (features, risk/protector) factors. The other is to develop new or adapt existing algorithms to analyze data from different sources.

Important Date

Regular Research Papers due                           June 15, 2008   July 15, 2008

Notification of acceptance                                           September 1, 2008

Camera-ready papers & Pre-registration                      October 1, 2008

The ICMLA Conference                                              December 11-13, 2008

Topics

Original research papers in the theory in data mining and machine learning with emphasis of applications in cancer genetics, cancer bioinformatics, and cancer epidemiology are solicited.  Specific topics include but not limited to:

Feature Selection and Biomarker Evaluations

• Feature ranking  and rank learning
• LASSO related feature selection
• Global optimization (genetic algorithm, MCMC) related feature selection
• Feature selection in clustering.
• Selection for labeled and unlabeled data
• Feature selection with epigenetic,  genomic and  proteomics data
• Wrapper methods for feature selection
• Bayesian feature selection 
• ROC based feature selection
• Selection bias
• Selection with small samples
• Cross-discipline comparative studies
• Feature selection with binary and time-to-event outcomes
• Optimal cutpoints and variable discretization methods
• ROC and Bayesian based methods for biomarker evaluations
• Comparative studies in different biomarker screening methods.

Data Mining and Machine Learning in SNP Tagging and Genomewide Association Studies

• Entropy and information theory related SNP tagging
• Data mining and SNP tagging with haplotype data
• Machine learning in genetic mapping with binary and quantitative traits
• Clinical outcome prediction with genotype data
•  Data mining and machine learning in genome-wide association study
• Machine learning to detect gene-gene interactions

Mining Data from Different Sources

• Algorithms to mine microarray and SNP data simultaneously
• Machine learning methods to mine gene expression and proteomic data together
• Gene network construction with data from different sources
• Spatial and temple data mining for cancer epidemiology data
• Efficient algorithms in spatial and temple data mining

Paper Submission:

Please submit a full length paper through the online submission system. Electronic submission is required. Authors will be notified of the acceptance after the review process by two independent reviewers. All papers accepted will be included in the Workshop Proceedings published by the IEEE Computer Society Press and will be available at the workshop.

If you have further questions, please contact the program chairs zliu@umm.edu or ysong@umes.edu.

Organization:

Special Session Chairs:

Zhenqiu Liu:         zliu@umm.edu

University of Maryland School of Medicine, USA

Yinglei Song: ysong@umes.edu

University of Maryland Eastern Shore, USA

Program Committee:

Halima Bensmail, University of Tennessee, Knoxville, USA

 Dong Xu, University of Missouri, USA

 Shili Lin, The Ohio State University, USA

  Dechang Chen, Uniformed Services University of the Health Sciences, USA

Haomiao Jia, Columbia University, USA.

 Ming Tan, University of Maryland, USA

Babis Papachristou, University of Chicago, USA      

Yunhu Wan, University of Maryland,  USA

JianJun (Paul) Tian, College of William and Mary, USA

Li Sheng, Drexel University, USA   

Wei Guo, The Ohio State University, USA

Russell L. Malmberg, University of Georgia, USA

Mark Kon, Boston University, USA

Chunmei Liu, Howard University, USA

Junfeng Qu, Clayton State University, USA