Organised by Neil Lawrence, Marc Dymetman.
For the launch talk detailing the theme for this programme see below.
TP1 Leveraging Complex Prior Knowledge for Learning
Neil D. Lawrence
Traditionally machine learning has focused mainly on constructing models in a data driven manner. Clearly, in practise, if we can incorporate domain knowledge with our learning we should be able to obtain improved performance. This type of knowledge is particularly important in application domains where data availability is sparse in the context of the complexity of the required model. In this thematic programme we will highlight and drive forward approaches to incorporating prior knowledge in the application domain. We are interested in all approaches to incorporating this prior knowledge and any application area. Already some subthemes (and application areas) are emerging within the programme for example: knowledge encoded in graph structures (applications in language and computational biology), knowledge encoded in ordinary and stochastic differential equations (applications in climate and systems biology) and knowledge encoded as probabilities (applications in language).
The Thematic Programme ran between March and September 2008.
Summer School in the Programme
- Modelling Biological Networks, Helsinki, Finland, 20th-24th October 2008
Workshops in the Programme
- Bayesian Approaches to Prior Knowledge
- Bayesian Research Kitchen, Grasmere, Lake District, U.K., 6th-7th September 2008
- Nonparametric Bayes Workshop, Helsinki, 9th July 2008 [PASCAL Video Lectures]
- Prior Knowledge encoded as Graphs
- Mining and Learning with Graphs, Helsinki, July 4th-5th 2008, [PASCAL Video Lectures]
- Prior Knowledge encoded in Stochastic/Dynamic Systems
- Approximate Inference in Stochastic Processes and Dynamical Systems, Cumberland Lodge, May 27th-29th 2008, [PASCAL Video Lectures]
- Prior Knowledge in Language
- Prior Knowledge for Text and Language Processing, Helsinki, 9th July 2008 [PASCAL Video Lectures]
- Biological Data and Prior Knowledge
- Learning and Inference in Computational Systems Biology, Glasgow, March 26th-27th, 2008
- Machine Learning in Systems Biology, Belgium, September 13th-14th, 2008
Challenge in the Programme
The GREAT08 challenge, organised by Sarah Bridle, is about estimating the level of gravitational lensing in images of galaxies. This challenging task requires prior knowledge about the effect of gravitational lensing (which in turn informs us about how much Dark Matter there is) and the effect of a telescope’s optics.