Brown University -
Division of Biology and Medicine,
Center for Statistical Sciences Seminar
| |
Abstract: Distances are used extensively in many scientific fields. They are also used in statistics but their role has not been emphasized in the statistical literature. In this talk, we discuss a class of distance measures that is characterized by a simple quadratic structure. Although this structure is very simple, the class contains a very rich and interesting class of distances and presents the opportunity for "handcrafted" distances designed for special problems. We exemplify the importance of this class of distances by illustrating connections with von Mises expansions and offering an L(2) interpretation. Additional motivation for studying these distances is provided through connections of this work with work from the scientific field of machine learning and the study of problems that arise in biomedical informatics.
Brown University -
Division of Biology and Medicine,
Center for Statistical Sciences Seminar
Department of Biostatistics, University of North Carolina | |
Abstract: Variable selection is an important topic in the statistical sciences with broad applications to many substantive areas. Additional difficulty is caused by censoring, which occurs when subjects are not followed until the endpoint of interest has occurred so their failure times are unobserved. We propose a method based on the smoothly clipped absolute deviation (SCAD) penalty (Fan and Li, 2001) and the accelerated failure time model using the Buckley-James estimator. While the accelerated failure time model has been sufficiently well- studied, the asymptotic properties and pragmatic performance for using the SCAD penalty with the Buckly-James estimator are unknown. We evaluate the proposed variable selection procedure through large and small sample studies.
Brown University -
Division of Biology and Medicine,
Center for Statistical Sciences Seminar
Candidate for Assistant Professor (tenure track) in the Public Health Program/Department of Community Health | |
Using the Subject-Specific Threshold ROC Curve | |
Abstract: Recent scientific and technological innovations have produced an explosion of potential biomarkers which are being investigated for their use in disease screening and diagnosis. In evaluating these new markers, it is often necessary to account for covariates which are associated with the biomarker of interest. For example, age is strongly associated with prostate-specific antigen (PSA), a biomarker for prostate cancer, and the discriminatory accuracy of PSA may also vary with age. We propose the subject-specific threshold ROC (SST-ROC) as a covariate-adjusted measure of the diagnostic accuracy of the biomarker. The SST-ROC is the ROC curve for a rule which uses covariate-specific thresholds to define "test-positive". It can also be interpreted as a weighted average of the covariate-specific sensitivities, holding the covariate-specific specificities constant. We motivate consideration of the SST-ROC, propose non-parametric and semi-parametric estimators, provide asymptotic distribution theory for these estimators, and explore the implications for efficient study design.
<--- 2005 Index