Graduate Student Seminar

Speaker 1: John Zylstra

Title

Effects of Model Misspecification in Linear Regression under Multiply Imputed Synthetic Data

Abstract

Government agencies that collect survey data require tools that permit them to to publish this data while preserving the identity of respondents.  Synthetic data is one such tool that takes advantage of the existing framework for multiple imputation of missing data.  The utility of synthetic data depends on the coincidence of various models and we examine the consequences of the absence of this coincidence for the linear regression model.

Speaker 2: Sumaya Alzuhairy

Title

Stationary Stochastic Processes

Abstract

It is actually possible to use the spectral measure obtained from the covariance function to represent a weak stationary processes. This is possible despite the fact that it has no Fourier transform. We discuss the way that shows such process has sort of generalized spectral representation.