逢 甲 大 學 應 用 數 學 系
專 題 演 講
講 者: Shaw-Hwa Lo
(Professor of Statistics and Biostatistics,
Department of Statistics,Columbia University)
題 目: A method of Detecting influential
variables with application to
Sporadic Breast cancer data
日 期: 100年12月22日(四)下午14:00-15:00
講 者: Tomohiro Ando
(Associate Professor of Management
Science, Graduate School of Business
Administration, Keio University)
題 目: Predictive approach for model selection
on econometric models with factor
augmented predictors
日 期: 100年12月22日(四)下午15:00-16:00
地 點:理學412室
(歡 迎 參 加 . 敬 請 張 貼)
2011/12/22（四）
14:00~15:00
Shaw-Hwa Lo
Professor of Statistics and Biostatistics, Department of Statistics,
Columbia University
A method of Detecting influential variables with application to Sporadic Breast
cancer data
A current challenge to statisticians is to develop effective methods of finding
the useful information from the vast amounts of messy and noisy data available, most
of which are noninformative. We review a general computer intensive approach for
detecting which and how, of a large number of explanatory variables, have an
influence on a dependent variable Y. This approach is suited to detect influential
variables, where causal effects depend on the confluence of values of several variables.
Applying this approach to a subset of case-control sporadic breast cancer data, (from
the National (CGEMS) initiative), focusing on 18 breast cancer related genes with
304 SNPs, indicates that there are many interesting interactions that form 2-way and
3-way networks in which BRCA1 plays a dominant and central role. The apparent
interactions of BRCA1 with many other genes suggests the conjecture that BRCA1
serves as a protective gene, and that some mutations in it or in related genes may
prevent it from carrying out this protective function even if the patients are not
carriers of BRCA1 mutations. The method of analysis features the evaluation of the
effect of a gene by averaging the effects of the SNP’s covered by that gene. Marginal
methods which test one gene at a time fail to show any effect. That may be related to
the fact that each of these 18 genes adds very little to the risk of cancer. Analysis
which relates the ratio of interactions to the maximum of the first order effects
discovers significant gene pairs and triplets.
15:00~16:00
Tomohiro Ando
Associate Professor of Management Science, Graduate School of
Business Administration, Keio University
Predictive approach for model selection on econometric models with factor
augmented predictors
Econometric models with a factor augmented regressors provide a useful
approach to forecasting when there are many predictors available. A properly chosen
model can summarize relevant information about the variable of interest in a small
number of indices and, hence, achieve substantial dimension reduction in the
predictor space. The application of econometric models with a factor augmented
regressors often involves two steps. In the first step, the common factors (or indices)
are estimated via the principal component analysis. In the second step, the estimated
common factors are used in conjunction with other pre-specified variables to build a
model for forecasting. For successful applications of these models, selection of the
best model is critical. In this paper, we propose a model selection criterion for
generalized linear, quantile regression models that takes into account the uncertainty
in estimated common factors. Results of real data analysis and Monte Carlo
simulations demonstrate clearly that the proposed criterion performs well.