S3RI Seminar - A test for local white noise (and the absence of aliasing) in locally stationary wavelet time series, Professor Guy Nason (University of Bristol) Seminar
- Time:
- 14:00 - 15:00
- Date:
- 24 May 2018
- Venue:
- University of Southampton Building 58, Seminar Room 1067 Highfield Campus Southampton SO17 1BJ
For more information regarding this seminar, please email Dr Helen Ogden at H.E.Ogden@soton.ac.uk .
Event details
Aliasing is when a time series is sampled at a rate that is not high enough to capture all information present in the underlying process and, traditionally, impossible to detect. This talk develops a new test for local white noise, which also doubles as a test for the lack of aliasing in a locally stationary wavelet time series. We compare and contrast our new test with the aliasing test for stationary time series due to Hinich and co-authors. We show that the test is robust to mismatch of analysis and synthesis wavelet. We demonstrate the effectiveness of the test on some simulated examples and on an example from wind energy. This work is joint with Idris Eckley, Lancaster
Speaker information
Professor Guy Nason , University of Bristol . Non-stationary time series. Multiscale methods in statistics. Variance stabilization. Computational statistics. Networks Official Statistics, Economics, Epidemiology, Energy