|Date:||October 24, 2013 from 1:10 pm to 2:10 pm EDT|
|Location:||Room 222 at Pupin Hall|
|Contact:||For further information regarding this event, please contact Jing by sending email to email@example.com .|
|Info:||Click Here to Visit Website.|
Speaker: Joseph Glaz, University of Connecticut
Title: Scan Statistics for Normal Data with Applications
Abstract: In this lecture probability inequalities and approximations for moving sums of iid normal random variables will be discussed. A connection with scan statistics for detecting a change in the mean process for the observed data will be presented. Both the case of variance known and unknown will be discussed. Extension of the results to ARMA time series models will be presented as well. R algorithms and simulations results will be presented as well as analysis of a data set for time series models. The lecture will be concluded with directions for future work and open problem.