Showing 1–3 of 3 results for author: Ambler, G K

Perfect simulation using dominated coupling from the past with application to area-interaction point processes and wavelet thresholding

Authors: Graeme K. Ambler, Bernard W. Silverman

Abstract: We consider perfect simulation algorithms for locally stable point processes based on dominated coupling from the past, and apply these methods in two different contexts. A new version of the algorithm is developed which is feasible for processes which are neither purely attractive nor purely repulsive. Such processes include multiscale area-interaction processes, which are capable of modelling… ▽ More We consider perfect simulation algorithms for locally stable point processes based on dominated coupling from the past, and apply these methods in two different contexts. A new version of the algorithm is developed which is feasible for processes which are neither purely attractive nor purely repulsive. Such processes include multiscale area-interaction processes, which are capable of modelling point patterns whose clustering structure varies across scales. The other topic considered is nonparametric regression using wavelets, where we use a suitable area-interaction process on the discrete space of indices of wavelet coefficients to model the notion that if one wavelet coefficient is non-zero then it is more likely that neighbouring coefficients will be also. A method based on perfect simulation within this model shows promising results compared to the standard methods which threshold coefficients independently. △ Less

Submitted 28 February, 2010; originally announced March 2010.

Comments: 27 pages, 8 figures. Chapter 3 of "Probability and Mathematical Genetics: Papers in Honour of Sir John Kingman" (Editors N.H. Bingham and C.M. Goldie), Cambridge University Press, 2010

Perfect simulation for Bayesian wavelet thresholding with correlated coefficients

Authors: Graeme K. Ambler, Bernard W. Silverman

Abstract: We introduce a new method of Bayesian wavelet shrinkage for reconstructing a signal when we observe a noisy version. Rather than making the common assumption that the wavelet coefficients of the signal are independent, we allow for the possibility that they are locally correlated in both location (time) and scale (frequency). This leads us to a prior structure which is analytically intractable,… ▽ More We introduce a new method of Bayesian wavelet shrinkage for reconstructing a signal when we observe a noisy version. Rather than making the common assumption that the wavelet coefficients of the signal are independent, we allow for the possibility that they are locally correlated in both location (time) and scale (frequency). This leads us to a prior structure which is analytically intractable, but it is possible to draw independent samples from a close approximation to the posterior distribution by an approach based on Coupling From The Past. △ Less

Submitted 15 March, 2009; originally announced March 2009.

Comments: 16 pages, originally Tech Report, University of Bristol, 2004

Perfect simulation of spatial point processes using dominated coupling from the past with application to a multiscale area-interaction point process

Authors: Graeme K. Ambler, Bernard W. Silverman

Abstract: We consider perfect simulation algorithms for locally stable point processes based on dominated coupling from the past. A version of the algorithm is developed which is feasible for processes which are neither purely attractive nor purely repulsive. Such processes include multiscale area-interaction processes, which are capable of modelling point patterns whose clustering structure varies across… ▽ More We consider perfect simulation algorithms for locally stable point processes based on dominated coupling from the past. A version of the algorithm is developed which is feasible for processes which are neither purely attractive nor purely repulsive. Such processes include multiscale area-interaction processes, which are capable of modelling point patterns whose clustering structure varies across scales. We prove correctness of the algorithm and existence of these processes. An application to the redwood seedlings data is discussed. △ Less

Submitted 15 March, 2009; originally announced March 2009.

Comments: 15 pages, originally a Bristol University 2004 Tech Report