Graduate Seminars

back Back to all Graduate Seminars

AMCS Graduate Seminar: Stationary subspace analysis of nonstationary covariance processes using matrix eigenstructures

Start Date: February 21, 2019
End Date: February 21, 2019

Dr. Raanju R. Sundararajan (KAUST)
Stationary subspace analysis (SSA) searches for orthogonal linear combinations of the components of nonstationary multivariate time series that are stationary. These linear combinations and their number define an associated stationary subspace and its dimension. SSA is studied here for zero mean nonstationary covariance processes. Basic linear algebra is applied to characterize stationary subspaces and their dimensions for these processes in terms of eigenvalues and eigenvectors of certain symmetric matrices. This characterization is then used to derive formal statistical tests for estimating dimensions of stationary subspaces. The asymptotic properties of these dimension tests are described. Eigenstructure-based techniques are also proposed to estimate stationary subspaces, without relying on previously used computationally intensive optimization-based methods. A graph-based clustering approach is adopted here to form clusters of the many subspaces. Finally, the introduced methodologies are examined on simulated and real data.
Bio: Dr. Raanju R. Sundararajan is a post-doctoral fellow at the Biostatistics group at KAUST. He completed his PhD in Statistics at Texas A&M University (College Station, Texas, USA) in 2018. His research focuses on time series analysis with an emphasis on handling time series data from nonstationary systems. He also works on applications of time series methods in areas such as neuroimaging and economics.

More Information:

For more info contact: Dr. Raanju Sundararajan : email:
Date: Thursday 21st Feb 2019
Time:12:00 PM - 01:00 PM
Location: Building 9, Lecture Hall 1 Room 2322
Refreshments: Light Lunch will be served at 11:45 AM