CAM Seminar
Spring 2023
February 10
Speaker: Dr. Mohamed Nasser
Title: The integral equation with the generalized Neumann kernel on circular domains
Abstract: This talk is about solving the boundary integral equation with the generalized Neumann kernel on circular domains using a method based on Fourier series. The method reduces solving the integral equation to solve a linear system with a matrix of special structure.
February 17 (Tom and Mai will meet bank people during seminar time)
February 24 (invited by Dr. Mai Dao)
Speaker: Dr. Nadeesha Jayaweera (WPI)
Title: Confidence Envelopes for Parametric Model Selection Criteria and Post-Model Selection Inference
Abstract: In choosing a candidate model in likelihood-based modeling by minimizing an information criterion, the practitioner is often faced with the difficult task of deciding how far up the ranked list to look. Motivated by this pragmatic necessity, we derive an approximation to the quantiles of a generalized (model selection) information criterion (GIC), defined as a criterion for which the limit in probability is identical to that of the normalized log-likelihood, and which includes common special cases such as AIC and BIC. The method starts from the joint asymptotic normality of the GIC values, and proceeds by deriving the (asymptotically) exact distribution of the minimum, which can be efficiently (numerically) computed. Inversion of this distribution function then provides the desired quantiles. The joint asymptotic of the GICs is derived in three cases of classical interest: (i) independent and identically distributed data, (ii) regression, and (iii) time series. The development in the latter two cases invokes Lindeberg-Feller type conditions for, respectively, normalized sums of the conditional distributions of the responses, and normalized quadratic forms in the observations. The performance of the methodology is tested on simulated data by assessing the nominal coverage probability of quantiles and com- pared to the bootstrap. Both approaches give coverages close to nominal for large samples, but the bootstrap is on average two orders of magnitude slower. Finally, we hint at the possibility of producing confidence intervals for individual parameters by pivoting the distribution of the minimum GIC, thus naturally accounting for post-model selection uncertainty.
March 3: No speaker
March 10: No speaker
March 17: Spring break
March 24 (invited by Dr. Mai Dao)
Speaker: Dr. Yet Nguyen (ODU)
Title: Variable selection in RNA-seq analysis by pseudo-covariate augmentation
Abstract: This talk presents a variable selection strategy in RNA-seq analysis using pseudo-covariates aiming at controlling the false selection rate, which is the expected proportion of selected irrelevant covariates. We show that the proposed method can accurately choose the most relevant covariates while holding the false selection rate below a specified level. Simulation studies demonstrate the advantages of our method over typical existing methods for detecting differentially expressed genes when covariates are available.
March 31: No speaker
April 7: No speaker
April 14: No speaker
April 21: No speaker
April 28: No speaker