Francesca Chiaromonte

Assistant Professor of Statistics
Ph.D., University of Minnesota, 1996

Summary of research interests

Dr. Chiaromonte's research concerns methods providing simplified representations of high-dimensional problems, with a particular focus on dimension reduction and graphics.

Some of Dr. Chiaromonte's work deals with general multivariate settings: She has developed a reduction framework to specify definitions of multivariate structure, and identify simplified representations preserving structural content which might live in lower dimension. Another part of her work is in the field of sufficient dimension reduction for regression, where variables play different roles as responses and predictors. Here, one identifies reduced regressions that are equivalent to the original one in terms of information on the response.

Dr. Chiaromonte is also interested in micro-founded (or agent-based) models, and in the statistical issues related to their exploration via simulation experiments and their calibration on real data. She has designed two such models within the economics domain (aggregate and structural effects of technical innovation; functioning of decentralized financial markets).

Recently she has become involved in genetics applications. In collaboration with several life science researchers at PSU and elsewhere, she is investigating relationships among local features of the human genome (divergence measures, GC content, repeat location, etc.), and analyzing global gene expression data from microarrays. These applications offer a challenging ground for dimension reduction methods, as well as clustering methods, which she is investigating from the perspective of mixture models.

Chiaromonte F., Cook R.D., Li B. (2001). Sufficient dimension reduction in regressions with categorical predictors. Annals of Statistics. To appear.

Chiaromonte F., Yang S., Elnitski L., Bing Yap V., Miller W., Hardison R.C. (2001). Association between divergence and interspersed repeats in mammalian noncoding genomic DNA. Proceedings of the National Academy of Sciences. To appear.

Chiaromonte F., Bing Yap V., Miller W. (2001) Scoring pairwise genomic sequence alignments. . To appear. Proceedings of the Pacific Symposium on Biocomputing 2002

Chiaromonte F., Martinelli J.A. (2001) Dimension reduction strategies for analyzing global gene expression data with a response. Mathematical Biosciences. To appear.

Chiaromonte F. (1998). On multivariate structures and exhaustive reductions. Computing Science and Statistics, 30, S. Weisberg (ed), Interface Foundation of North America, Inc. Fairfax Station VA, 204-213.

Chiaromonte F. Dosi G. (1998). Modeling a decentralized asset market: an introduction to the financial "Toy-Room". Interim Rep. IR-98-081, IIASA, Laxenburg AUSTRIA.

Chiaromonte F. (1997). A reduction paradigm for multivariate laws. L1 Statistical Procedures and Related Topics, Y. Dodge (ed), Institute of Mathematical Statistics Monograph Series, Hayward CA, 229-240.

Chiaromonte F. Cook R.D. (1997). On foundations of regression graphics. Tech. Rep. 616, School of Statistics, University of Minnesota, 1997

Chiaromonte F. Dosi G. (1993). Heterogeneity, competition, and macro-economic dynamics. Structural Change and Economic Dynamics 4,1, 39-63.