Candidate Statements

Rebecca Nugent

Present Position

Visiting Assistant Professor (NSF VIGRE Postdoctoral Fellow), Department of Statistics, Carnegie Mellon University (Sept 2006 - present)

Former Positions:

Pre-doctoral Lecturer, Department of Statistics, University of Washington (Jan 2006 - May 2006)

Degrees

• PhD, Statistics, University of Washington, 2006
• MS, Statistics, Stanford University, 2001
• BA, Mathematics, Statistics, Spanish (Language/Linguistics), Rice University, 1999

Field(s) of major professional activity:
nonparametric clustering/classification; graphics and visualization; uncertainty in clustering
public health applications: analysis of sleep patterns and their stability; robustness of pulmonary function testing

Publications (list up to three major articles or books)

Stuetzle, W. Nugent, R. "A generalized single linkage method for estimating the cluster tree of a density". Journal of Computational and Graphical Statistics. In revision.

Buscemi, D., Kumar, A., Nugent, R., Nugent, K. "Short Sleep Times Predict Obesity in Internal Medicine Clinic Patients". Journal of Clinical Sleep Medicine (in press).

Other journals published in, or books:

Cancer Causes Control, American Journal of Epidemiology, International Journal of Cancer

Professional activities (editorships, offices held, etc., if any)
Reviewer: Statistics in Medicine, JRSS-B, Psychometrika

Chair of contributed sessions at Joint Statistical Meetings, August 2006 and 2007

Other relevant information (you may include a statement of your view of CSNA)

CV at http://www.stat.cmu.edu/~rnugent/CV.pdf

In my view, the strengths of CSNA lie in its tight-knit community and the specialized focus of its conferences. CSNA and its members provide a supportive environment that allows new researchers, new faculty, and graduate students to debate research ideas and develop as statisticians while learning from experts in the field. The specialized focus of the meetings allows for immersion in the current clustering/classification research that larger conferences could never provide. I would like to develop outreach programs that emphasize these qualities that should be particularly attractive to new statisticians. These programs could include, for example, increased advertising at other conferences or developing a young researchers invited session short-paper competition (for either new faculty or graduate students). In addition, the increase in interdisciplinary centers for applied statistics at universities world-wide should be an excellent opportunity to supplement the sessions of methodological talks with sessions focusing on specific applications of clustering and classification. These sessions might be attractive to those pursuing applied careers and could potentially increase both attendance and the participation of other related fields. My experience with CSNA as a graduate student and then as a young faculty member has been exceedingly positive; my focus will be on continuing to maintain the closeness of the community while extending invitations to new members in both statistics and related disciplines.

Bill Shannon

Present Position

Associate Professor of Biostatistics in Medicine, Washington University School of Medicine, St. Louis, MO USA

Degrees

• MS in Zoology, Univ. of Massachusetts
• PhD in Biostatistics, Univ. of Pittsburgh

Field(s) of major professional activity:

• Applied biostatistics in medicine
• Cluster analysis with applications to genetics
• Statistical computing
• Statistical consulting

Publications (list up to three major articles or books)

Shannon, W. Cluster Analysis, in Handbook of Statistics, Vol. 27, eds. CR Rao, DC Rao, Phil Miller (in press).

Shannon, W. 'Biostatistical Challenges in molecular data analysis,' In Discrete methods in Epidemiology: DIMACS Workshop on Data Mining and Epidemiology, eds. James Abello, Graham Cormode, Vol. 70. 2007.

Shannon W, Culverhouse R, Duncan J. Analyzing microarray data using cluster analysis. Pharmacogenomics 2003; 4(1):41-52.

Professional activities

• Board of Directors CSNA
• Board of Directors Interface
• Organizer of past CSNA and Interface meetings

Other relevant information

CSNA is in a membership crisis. Without a quick and significant increase in members we face the likelihood of folding as an organization. This is in many ways inconceivable to me considering how cluster and classification is used in so many disciplines. My focus as President-elect and President will be on coming up with ways of increasing and diversifying membership working with members and the board to develop a new focus in CSNA that will attract a broader membership class. CSNA was and should be again the focal point for anyone in any discipline doing applied, methodology, and/or theoretical work in cluster and classification.

Douglas Steinley

Present Position:

Assitant Professor, Psychological Sciences, University of Missouri

Degrees:

• Ph.D. (Quantitative Psychology). 2004. University of Illinois
• M.S. (Statistics). 2002. University of Illinois
• M.A. (Psychology). 2001. University of Illinois
• B.Sc. (Statistics). 2000. University of Missouri
• B.A. (Psychology). 2000. University of Missouri.

Fields of major professional activity: Multivariate statistics, cluster analysis, combinatorial data analysis

Publications:

Steinley, D. (2007). Validating clusters with the lower bound for sum of squares error. Psychometrika, 71, 93--106.

Steinley, D. (2006). Profiling local optima in K-means clustering: Developing a diagnostic technique. Psychological Methods, 11, 178--192.

Steinley, D. (2006). K-means clustering: A half-century synthesis. British Journal of Mathematical and Statistical Psychology. 59, 1--34.

Other journals published in: Multivariate Behavioral Research, Journal of Classification, IIE Transactions, The Gerentologist, Journal of Mathematical Psychology, Journal of Aging and Health, Journal of the American Academy of Child & Adolescent Psychiatry

Statement:

To increase membership in CSNA from the behavioral and social sciences, it is important that procedures are shown to be relevant to commonly asked problems. To date, my research agenda has been to show how many of the more exploratory techniques used by CSNA members are at least as relevant as--if not preferable-- to more comman analytic strategies. I hope to continue to publish research in highly-visible quantitative psychology outlets that promotes those goals.