A photo from one of the workshopsSmoothing-Based and Gaussian-Process-Based Methods for Non-Parametric Regression in Environmental Problems

Mar 26, 2007 - Mar 30, 2007

ICMS, 14 India Street, Edinburgh

Organisers

Name Institution
Anderson, Clive University of Sheffield
Barnett, Vic Nottingham Trent University
Bowman, Adrian University of Glasgow
Scott, Marian University of Glasgow
Smith, Richard University of North Carolina
Smith, Ron CEH Edinburgh
Turkman, Antonia University of Lisbon

Short Report

The workshop aimed to provide an overview of the current state of research in smoothing and Gaussian process based approaches to non-parametric regression with applications in environmental sciences.   A brief summary of the content is as follows: six speakers gave a series of presentations that reflected new developments in Bayesian and Gaussian process modeling, four speakers outlined the challenges in environmental science; three talks discussed connections between the three main areas of the workshop; and six speakers addressed new developments and applications. For example, Professor Marian Scott and Vic Barnett opened the workshop with a brief introduction to SPRUCE (co scientific sponsor) and to the themes of the workshops, these being then followed by more technical introductions to smoothing, Gaussian process and the nature of environmental problems.  Mark Hallard (SEPA), Gavin Simpson (UCL- Environmental change) and Jan Dick and Peter Levy (CEH) all spoke about their needs as environmental scientists.

Download the pdf file of the full report 

 Original Details

The meeting will bring together leading experts in (a) modern smoothing techniques in data analysis; (b) Gaussian process based Bayesian statistical methodology; and (c) environmental statistics. The objectives will be to:

  • explore the relationship between smoothing-based and Bayesian-Gaussian-process-based approaches to non-parametric regression and model uncertainty, and assess the potential for transfer of computational tools and theoretical constructs between the two;
  • confront non-parametric regression methodology with large scale environmental problems of trend detection and assessment, and of model uncertainty, and to formulate proposals for practical solutions;
  • explore the power of non-parametric regression techniques in spatial and spatio-temporal contexts and consider advantages and disadvantages of the techniques with any computational issues.

Discussion themes include:

Theme 1: Challenges in modern smoothing techniques in regression and time series: model-based opportunities?

Theme 2: Gaussian processes and Bayesian methodology: strategies for implementation in high-dimensional environmental contexts.

Theme 3: Environmental science problems and questions needing a statistical solution. Too much data, too little information?

Theme 4: Spatial and spatio-temporal non-parametric modelling

Theme 5: Computational challenges


Numbers will be limited to a maximum of 40, there will be a contribution to local costs (including accommodation and subsistence) from an award to ICMS from the Engineering and Physical Sciences Research Council and a Registration Fee of £30 will be payable.


Those interested in attending should contact Marian Scott (marian@stats.gla.ac.uk) or Clive Anderson (c.w.anderson@shef.ac.uk) by January 15 2007.

Arrangements

Venue
The Workshop will be held at the head-quarters of ICMS at 14 India Street. This house is the birthplace of James Clerk Maxwell and is situated in the historic New Town of Edinburgh, very near the centre. The seminar room is fully equipped with large whiteboards, overhead and data projectors. Other facilities include PCs, wireless network and rooms suitable for small discussion meetings.

The ICMS travel pages show the location of India St and have links to pages with advice on how to get to Edinburgh and travel around once here.

Registration etc
Registration will be from 12.00-14.00 on Monday 26, with a sandwich lunch provided. The talks will start at 14.00 and the workshop will end at 12.00 on Friday 30.

Accommodation
ICMS will arrange rooms in local guest houses for those who require it. Participants are also free to make their own arrangements and may claim back the cost, with receipts, up to a maximum of £45 per night for bed and breakfast. A list of Edinburgh accommodation of various sorts and prices is available here. Sections 1-3 are particularly relevant.

Meals and Refreshments
Morning and afternoon refreshments will be provided throughout the Workshop. A sandwich lunch will be supplied on Monday only. There will be a wine reception on Monday evening, a group meal on Tuesday evening and the Workshop Dinner on Thursday evening.

Further information will be provided in the Invitation Letter.

Programme


Monday 26 March
12.00-14.00 Sandwich lunch & Registration
14.00-15.30 Introduction sessions
Vic Barnett
Marian Scott
Round table and introductions, discussion groups set
15.30-16.00 Coffee
16.00-17.30 The workshop themes
Clive Anderson
tba
Adrian Bowman
The use of additive models in environmental modelling
Marian Scott
What makes some environmental problems especially challenging statistically?
17.30 Reception with wine

Tuesday 27 March
09.00-10.30 Gaussian and Bayesian based process models
Nial Friel
Inference for binary Markov random fields without simulation
Dan Cornford
Data assimilation, Bayesian inference and Gaussian processes: opportunities and challenges
10.30-11.00 Coffee
11.00-12.30 Gaussian and Bayesian based process models
Mark Girolami
tba
Mannfred Opper
A variational Gaussian approximation for stochastic differential equations
12.30-14.00 Lunch break
14.00-15.30 Environmental science issues and needs
Mark Hallard
tba
Gavin Simpson
Identifying patterns of change in noisy water chemistry time series data
CEH speakers - Janet Dick, Peter Levy
15.30-16.00 Coffee
16.00-17.30 Environmental science issues and needs
Discussion
19.30 Group meal at Nargile, Hanover Street

Wednesday 28 March
09.00-10.30 Gaussian and Bayesian based processes
Chris Williams
Gaussian processes in machine learning
Havard Rue
Approximate Bayesian inference for latent Gaussian models using integrated nested Laplace approximations
10.30-11.00 Coffee
11.00-12.30 Discussion groups
12.30-14.00 Lunch break
14.00-15.30

PhD and postdoc session
Short talks from the PhD and post docs

Boukouvalas' presentation

Butler's presentation

15.30-16.00 Coffee
16.00-17.30 Smoothing and non-parametric models
Adrian Bowman
The use of additive models in environmental modelling
John Haslett
Monotone smoothing: application of a compound Poisson-Gamma process to modelling radiocarbon-dated depth chronologies

Thursday 29 March
09.00-10.30 General applications
Feridun Turkman
Spatial extremes of wild-ï¬re sizes; A Bayesian Hierarchical model
Anthony Davison
Smoothing and Temperature Extremes
Download presentation
10.30-11.00 Coffee
11.00-12.30 General applications
12.30-14.00 Lunch break
14.00-15.30 Uncertainties
Peter Challenor
Thermohaline Bayesian analysis
Download presentation
John Paul Gosling
Uncertainty in environmental modelling: carbon flux calculations for England and Wales
15.30-16.00 Coffee
16.00-17.30 Round-up discussion
19.30 Workshop Dinner at Howies, Waterloo Place

 

Participants

Name Institution
Anderson, Clive University of Sheffield
Barnett, Vic Nottingham Trent University
Boukouvalas, Alexis Aston University
Bowman, Adrian University of Glasgow
Butler, Adam Biomathematics and Statistics Scotland
Challenor, Peter National Oceanography Centre
Cornford, Dan Aston University
Crujeiras, Rosa University of Santiago de Compostela
Csato, Lehel Babes-Bolyai University
Daneshkhah, Alireza University of Strathclyde
Davison, Anthony École Polytechnique Fédérale de Lausanne
Dick, Janet CEH Edinburgh
Ferguson, Claire University of Glasgow
Friel, Nial University of Glasgow
Gattiker, James University of Southampton
Girolami, Mark University of Glasgow
Gosling, John Paul University of Sheffield
Haslett, John Trinity College Dublin
Holmes, Chris University of Oxford
Illian, Janine University of St Andrews
Lee, Dae-Jin Universidad Carlos III de Madrid
Levy, Peter CEH Edinburgh
Marrel, Amandine French Atomic Energy Commission
Opper, Manfred Technische Universität Berlin
Pan, Jianxin University of Manchester
Parnell, Andrew Trinity College Dublin
Retkute, Renata University of Surrey
Rue, Havard Norwegian University of Science & Technology
Scott, Marian University of Glasgow
Simpson, Gavin University College London
Smith, Ron CEH Edinburgh
Turkman, Antonia University of Lisbon
Turkman, Feridun University of Lisbon
Vrettas, Michail Aston University, Neural Computing Research Group (NCRG)
Williams, Chris University of Edinburgh