Personal tools

A new approach to dimension reduction for multivariate time series

— filed under:

Chung Eun Lee, University of Illinois at Urbana-Champaign

  • Colloquium
When Mon, Feb 20, 2017
from 11:00 AM to 11:50 AM
Where Ritter Hall 323
Add event to calendar vCal

Abstract: In this talk, we introduce a new methodology to reduce the number of parameters in multivariate time series modeling. Our method is motivated from the consideration of optimal prediction and focuses on the reduction of the effective dimension in conditional mean of time series given the past information. In particular, we seek a contemporaneous linear transformation such that the transformed time series has two parts with one part being conditionally mean independent of the past information. Our dimension reduction procedure is based on eigen-decomposition of the so-called cumulative martingale difference divergence matrix, which encodes the number and form of linear combinations that are conditional mean independent of the past. Interestingly, there is a factor model representation for our dimension reduction framework and our method can be further extended to reduce the dimension of volatility matrix. We provide a simple way of estimating the number of factors and factor loading space, and obtain some theoretical results about the estimators. The new method is applied to GDP and 7-city temperature series to illustrate the usefulness of the new approach.

« April 2017 »
Upcoming Events
PhD Oral Defense
Thu, Apr 27, 2017
Realizing Injective Splittings of Stable 4-Manifolds Gerrit Smith, SLU
Geometry/ Topology Seminar
Tue, May 02, 2017
The Causal Boundary Construction for Spacetimes, 11 Stacey Harris, SLU
Fri, May 05, 2017
Gilbreath Knots Jacqueline Jensen-Vallin, Lamar University
Annual Awards Ceremony
Fri, May 05, 2017
Let’s Get Knotty! Jacqueline Jensen-Vallin, Lamar University
Previous events…
Upcoming events…