MULTL-FREQUENCY SCALE GAUSSIAN REGRESSION FOR NOISY TIME-SERIES DATA
Author(s): Kian Seng Neo, W. E. Leithead
Type: Conference Paper (Proc. of the 6th bi-ennial UKACC Control Conference, International Control Conference 2006)
Date: August 2006
Abstract: Regression using Gaussian process models is applied to time-series data analysis. To extract from the data separate components with different frequency scales, the Gaussian regression methodology is extended through the use of multiple Gaussian process models. Fast and memory-efficient methods, as required by Gaussian regression to cater for large time-series data sets, are discussed. These methods are based on the generalised Schur algorithm and a procedure to determine the Schur decomposition of matrices, the key step to realising them, is presented. In addition, a procedure to appropriately initialise the Gaussian process model training is presented. The utility of the procedures is illustrated by application of a multiple Gaussian process model to extract separate components with different frequency scales from a 5000-point time-series data set with gaps.
Conferences
- Leithead, W. E., Neo, Kian Seng and Leith, D. J., Gaussian regression based on models with two stochastic processes, 16th IFAC World Congress, Prague, July 2005.
- Leithead, W. E., Zhang, Yunong and Neo, Kian Seng, Wind turbine rotor acceleration: identification using Gaussian regression, 2nd ICINCO, Barcelona, September 2005.
- Neo, Kian Seng, Leithead, W. E., Zhang, Yunong, Multi-frequency scale Gaussian regression for noisy time-series data, 6th bi-ennial UKACC Control Conference, International Control Conference 2006, Glasgow, September 2006. (Accepted)
Keith Web Logo 