steele@math.mcgill.ca |
|
Dept.
of Mathematics and Statistics |
350
Prince Arthur O. Apt. D727 |
|
McGill
University |
Montreal,
QC H2X 3R4 |
|
805
Sherbrooke West |
(514)
287-1388 |
|
Montreal,
QC H3A 2K6 |
|
|
(514) 398-3837 |
Education |
University of Washington Fall 1998 – August 2002 Ph.D.
in Statistics Carnegie Mellon University Fall 1993 – August 1998 M.
S. in Statistics August
1998 B.
S. in Statistics, May 1997 Minor
in Information and Decision Systems |
Employment |
McGill University Fall
2002 – Present Assistant Professor of Statistics
Courses: Fall 2002, Principles
of Statistics I (MATH 203). Winter 2003, Computation Intensive Statistics
(MATH 680). Research: Current
research includes Bayesian analysis of mixture models and numerical
integration approaches to multiple imputation. |
Publications |
Practical
Importance Sampling Methods for Finite Mixture Models and Multiple
Imputation, Steele, R.J. Unpublished
Ph.D. Thesis, University of Washington. (2002) Easy Computation
of Bayes Factors and Normalizing Constants for Mixture Models Via Mixture
Importance Sampling, Emond, M.J.,
Raftery, A.E., and Steele, R.J. University of Washington Technical Report
No. 398. (2001) Computing the Exact Distribution for a Multi-Way
Contingency Table Conditional on Its Marginal Totals, Fienberg, S.E.,
Makov, U.E., Meyer, M.M., and Steele, R.J. In A.K.M.E. Saleh, ed.,
Data Analysis from Statistical Foundations: A Festschrift in Honor of the
75th Birthday of D. A. S. Fraser, Nova Science Publishers, Huntington, NY,
145-165. (2001) Contribution to the Discussion of Stephens and Donnelly, Inference in Molecular Population Genetics, Emond, M.J., Raftery, A.E., and Steele, R.J. Journal of the Royal Statistical Society, Series B, 62. (2000) Disclosure Limitation Using Perturbation and Related Methods for Categorical Data, Fienberg, S.E., Makov, U.E., and Steele, R.J. Journal of Official Statistics, 14, No. 4. (1998) Statistical Notions of Data Disclosure Avoidance and Their Relationship to Traditional Statistical Methodology: Data Swapping and Loglinear Models, Fienberg, S.E., Steele, R.J., and Makov, U. 1996 Annual Research Conference of the U.S. Census Bureau. (1996) |
Research Interests |
Statistical computing, mixture models, multiple imputation, Bayesian modeling and inference, model-based clustering, large datasets, loglinear models, disclosure avoidance. |
Research Experience |
Doctoral Level Research Assistant: Studied the use of Bayes factors in determining the number of components in mixture and clustering models, developing sensible default priors for mixture models, and designing and choosing imputations for large-scale missing data problems. Advisor: Dr. Adrian Raftery, University of Washington Master’s Level Research Assistant: Conducted statistical analysis of patient eye movement data with Western Psychiatric Institute and Clinic, Pittsburgh. Advisor: Dr. Joel Greenhouse, Carnegie Mellon
University |
Invited Talks And
Selected Seminars |
Continuing Education Short Course on Model Based Clustering With Adrian Raftery, Chris Fraley, and John Castelloe ASA Joint Statistical Meetings, August 2001 Bayes Factors for Finite
Mixture Models from the EM Algorithm Via Importance Sampling Co-sponsored by the Section on Bayesian Research Methods and the Section on Survey Research Methods ASA Joint Statistical
Meetings, August 2000 |
Recognition and Awards |
Graduate Student Representative to faculty for Statistics
department, ARCS (Achievement Rewards for College Scientists) Fellowship
Recipient, Andrew Carnegie Society Presidential Scholar, Phi Beta Kappa and
Phi Kappa Phi Honor Societies |