skip to content

EC Summer School: Bayesian Signal Processing

19th July 1998 to 31st July 1998

Original URL:

Isaac Newton Institute for Mathematical Sciences

Nonlinear and Nonstationary Signal Processing

July to December 1998

Organisers: W J Fitzgerald (Cambridge), R L Smith (University of North Carolina), A Walden (Imperial College, London) and P C Young (Lancaster University)

EC Summer School

Bayesian Signal Processing

Organisers: WJ Fitzgerald (Cambridge), RL Smith (North Carolina),
AT Walden (Imperial), PC Young (Lancaster)

At the Newton Institute, Cambridge, UK

19 - 31 July 1998

The main focus and thrust of this workshop will be that Bayesian methods provide a unifying methodology whereby different kinds of mathematical models may be examined within a common statistical framework. The workshop will bring together the statistical and computational expertise of leading statisticians and the modelling expertise of mathematicians and subject matter specialists, with the broad objective of developing new signal processing tools which make efficient use of modern computational resourc es while combining the most up-to-date research of both groups of specialists.

The workshop will take place at the start of a six month programme on Nonlinear and Nonstationary Signal Processing to be held at the Isaac Newton Institute in Cambridge (July - Dec 1998), where it is hoped that many of the problems identified during the workshop will be studied in detail by the participants.

Final Programme (Updated 18.06.98)

Sunday 19 July

7pm - ...(please note change from 6pm) Reception at Christ's College

Monday 20 July

09.00 - 10.00 Registration

10.00-11.00 W J Fitzgerald (Cambridge)
An Introduction to Bayesian Signal Processing I

11.00-11.30 Coffee

11.30-12.30 W J Fitzgerald (Cambridge)
An Introduction to Bayesian Signal Processing II

12.30-14.00 Lunch

14.00-15.00 C Robert (Paris)
Latent Variable Models

15.00-15.30 Tea

15.30-16.30 C Robert (Paris)
The Simulated Likelihood Method

Tuesday 21 July

10.00-11.00 P J W Rayner (Cambridge)
Non-linear Signal Modelling - a Tutorial I

11.00-11.30 Coffee

11.30-12.30 P J W Rayner (Cambridge)
Non-linear Signal Modelling - a Tutorial II

12.30-14.00 Lunch

14.00-15.00 J A Stark (NISS and Cambridge)
MCMC Methods for Multiple Changepoint Inference

15.00-15.30 F Gustafsson (Sweden)
Automotive applications on adaptive filtering and change

15.30-16.00 Tea

16.00-17.00 J Scargle (NASA)
Multiple Change Point Determination for Point Processes

Wednesday 22 July

10.00-11.00 C Bishop (Microsoft Research)
Introduction to Probabilistic Graphical Models I

11.00-11.30 Coffee

11.30-12.30 C Bishop (Microsoft Research)
Introduction to Probabilistic Graphical Models II

12.30-14.00 Lunch

14.00-15.00 D Spiegelhalter (Cambridge)
Bayesian graphical modelling of complex applications

15.00-15.30 Tea

15.30-16.30 Introduction to 'BUGS'

16.45 -17.45 David Mackay (Cambridge)
Latent Variable Models and Neural Networks

6.00- ... Wine reception and Buffet (Sponsored by Cambridge Neurodynamics and Autonomy)

Thursday 23 July

9.00-10.00 M Niranjan (Cambridge)
Bayesian Techniques in Speech Signal Processing

10.00-11.00 David Mackay (Cambridge)
State of the art in error correcting codes

11.00-11.30 Coffee

11.30-12.30 S J Godsill (Cambridge)
Bayesian methods for non-Gaussian signal processing

12.30-14.00 Lunch

14.00-14.30 F Gustafsson (Sweden)
MDL as a performance measure for choosing adaptive filters.

14.30-16.00 M Pitt (Imperial College)
Filtering and Smoothing for Non-Gaussian state space models

16.00-16.30 Tea

16.30-17.30 John Skilling (Cambridge)
Annealing and Ensembles

Friday 24 July

10.00-11.00 Giulio D'agostini (Rome)
A Tutorial on Measurement Uncertainty I

11.00-11.30 Coffee

11.30-12.30 Giulio D'agostini (Rome)
A Tutorial on Measurement Uncertainty II

12.30-14.00 Lunch

14.00-15.00 Siddhartha Chib, Neil Shephard and Ola Elerian
Likelihood inference for discretely observed non-linear diffusions

15.00-16.00 V Krishnamurthy
Iterative Maximum A-posteriori Sequence Estimation for Nonlinear Systems

16.00-16.30 Tea

16.30-17.30 A Hyvarinen (Finland)
Independent Component Analysis, Blind Source Separation, and Projection Pursuit

19.30 Conference Dinner (Christ's College)

Saturday 25 July
Sunday 26 July         Punting etc.

Monday 27 July

9.00-10.00 Mike West (Duke)
Decomposition Methods In Nonstationary Signal Processing

10.00-11.00 A Walden (Imperial College)
Scale Analysis of Time Series

11.00-11.30 Coffee

11.30-12.30 R Baraniuk
Wavelet-domain statistical signal and image processing using hidden markov models

12.30-14.00 Lunch

Cricket Match and Barbeque

Tuesday 28 July

10.00-11.00 R Nowak
Bayesian Analysis of Poisson Processes: A Wavelet-Based Approach

11.00-11.30 Coffee

11.30-12.30 B Vidakovic (Duke)
Nonlinear Bayesian Shrinkage in the Wavelet Domain by Mixtures of Normal-Inverse Gamma Priors

12.30-14.00 Lunch

14.00-15.00 P Mueller (Duke)
Bayesian Inference with Wavelets: The Non-Equidistant Case - An MCMC Scheme

15.00-16.00 D Denison (Imperial College)
Nonparametric Bayesian models for regression

16.00-16.30 Tea

16.30-17.00 M Hild (Cambridge)
A Non-Probabilistic Measure of Plausibility

16.00 - 18.30 Posters

Wednesday 29 July

10.00 -11.00 N Shephard (Oxford)
An example of signal processing in finance: stochastic volatility

11.00-11.30 Coffee

11.30-12.30 P C Young (Lancaster)
Recursive estimation of time and state dependent parameters using fixed interval smoothing

12.30-14.00 Lunch

14.00-14.30 A Doucet (Cambridge)
MCMC Methods for Blind Bayesian Deconvolution

14.30-15.00 C Andrieu (Cambridge)
MCMC Methods for Bayesian Spectral Analysis

15.00-16.00 AP Quinn (TCD)
Bayesian perspectives on signal identification

16.00-16.30 Tea

16.30-17.30 E Moulines and O Cappe (Paris)
Estimation via simulation with applications to statistical inference in incomplete data models

Thursday 30 July

9.00 -11.00 R Smith
Spatial Statistics

11.00-11.30 Coffee

11.30-12.30 S Wilson (TCD, Dublin)
Bayesian Image Segmentation

12.30-14.00 Lunch

14.00-15.00 A Kokaram (TCD, Dublin)
Joint Detection and Removal of Missing Data for Degraded Image Sequences

15.00-16.00 A Hero
Robust Entropy Estimation via Pruned Minimal Spanning Trees

16.00-16.30 Tea

16.30-17.30 Roderick Murray-Smith (Denmark)
Multiple Model Approaches to Empirical Modelling

Friday 31 July

10.00 -11.00 P Djuric (Stony Brook)
On variable duration Hidden Markov Models

11.00-11.30 Coffee

11.30-12.30 R Elliott (Canada)
Kronecker's Lemma and parameter estimation

12.30-14.00 Lunch

14.00-15.00 J Scargle (NASA)
Exciting Events in the Universe

15.00-16.00 R Kohn (New South Wales)
Inference for dynamic mixture models

16.00-16.30 Tea

16.30-17.30 Leonard Smith (Oxford)
Shadowing and the dynamics of uncertainty within a given model class

17.30-18.00 Closing words


1) Albert C.S. Wong
On a Mixture Autoregressive Model

2) Maria Eduarda R. P. A. da Silva
Frequency Domain Estimation of Hidden Markov Autoregressive Models.

3) Dr. Gul Ergun
Bayesian Modelling of the Istanbul Stock Exchange.

4) Bart Mertens
Bayesian prediction with latent growth models for on-line monitoring in cheese manufacture.

6) James Heald
Use of the Evidence method to estimate appropriate measurement and dynamical noise levels for chaotic noise reduction

7) V. B. Mani Mohan and W. J. Fitzgerald
Bayesian estimation in the cyclic correlation domain

8) Jacek Noga and W. J. Fitzgerald
Hybrid Monte Carlo for State Space Models

9) Simon Barker
MCMC approaches applied to Image Segmentation.

10) Jasvinder Kandola
Statistical Modelling of Manufacturing Processes.

11) Colin Campbell and S J Godsill
Simulation for conditionally Gaussian time series using a stochastic version of EM

12) J A Stark, W J Fitzgerald and S B Hladky
MCMC Modelling of Ion Channels

13) B Kannan and W J Fitzgerald
A Numerical Bayesian Approach for DOA and Frequency Estimation of Exponential Signals in Gaussian and Non-Gaussian Noise.

14) JFG de Freitas
Sequential Monte Carlo Methods Applied to Neural Networks in Non-Linear, Non-Stationary Environments

15) Mike Davies
Blind Signal Separation using Independent AR models

16) T Clapp
Fixed-lag Smoothing using Sequential Importance Sampling

17) Jaco Vermaak
MCMC Estimation of a Speech Production Model for Voiced Sounds

18) David Leporini
Reversible jump MCMC methods for wavelet packet decompostion of transient signals

19) Paul T. Troughton and Simon J. Godsill
MCMC Methods for Restoration of Nonlinearly Distorted Autoregressive Signals

20) Bruno Sanso ISDS, Duke University
A Multisite Dynamic Model for Rainfall

21)Paul Walmsley, Simon Godsill and Peter Rayner
Multidimensional Optimisation of Harmonic Signals

University of Cambridge Research Councils UK
    Clay Mathematics Institute London Mathematical Society NM Rothschild and Sons