SPEAKER: Vladas Pipiras; Professor, Statistics and Operations Research, UNC Chapel Hill
TITLE: “Eigenstructure description and testing in stationary subspace analysis”
ABSTRACT: For data collected over time and multiple sources, a stationary subspace analysis (SSA) seeks linear combinations of data over different sources that are stationary in time (e.g. in the sense of constant mean and/or constant variance). The number of linearly independent combinations is referred to as the dimension of a stationary subspace. The talk will discuss new insights and new approaches to SSA by relating the quantities of interest in SSA, e.g. the stationary subspace dimension, to those from the eigenstructures of suitable matrices, and then estimating and testing for these through suitable matrix estimators. Some simulations and data applications will also be presented.
SPEAKER'S WEB PAGE: http://pipiras.web.unc.edu/
DATE: Thursday, February 21st, 4:10pm
LOCATION: MSB 1147, Colloquium Room
REFRESHMENTS: 3:30pm MSB 4110 (4th floor lounge)
STA 290 Seminar List: https://statistics.ucdavis.edu/seminars