Workshop on Soft Methods in Probability and Statistics

Lead Research Organisation: University of Bristol
Department Name: Engineering Mathematics

Abstract

The SMPS 2006 workshop will provide a forum for research and discussion into the fusion of soft methods with probability and statistics, with the ultimate goal of integrated uncertainty modelling in complex systems involving human factors. In addition to probabilistic factors such as measurement error and other random effects, the modelling process often requires us to make qualitative and subject judgements that cannot easily be translated into precise probability values. Such judgements give rise to a number of different types of uncertainty including; fuzziness if they are based on linguistic information; epistemic uncertainty when their reliability is in question; ignorance when they are insufficient to identify or restrict key modelling parameters; imprecision when parameters and probability distributions can only be estimated within certain bounds. Statistical theory has not traditionally been concerned with modelling uncertainty arising in this manner but soft methods, a range of powerful techniques developed within AI, attempt to address those problems where the encoding of subjective information is unavoidable. These are mathematically sound uncertainty modelling methodologies which are complementary to conventional statistics and probability theory. Therefore, a more realistic modelling process providing decision makers with an accurate reflection of the true current state of our knowledge (and ignorance) requires an integrated framework incorporating both probability theory, statistics and soft methods. This fusion motivates innovative research at the interface between computer science (AI), mathematics and systems engineering. NOTE: In view of the fact that soft methods form part of the AI research area we feel that it would be most appropriate if this proposal were initially considered by the ICT panel

Publications

10 25 50
 
Description The conference was attended by academics, researchers and PhD students from over 16 countries including Belgium, China, France, Germany, Holland, Israel, Italy, Japan, Poland, Slovakia, Solvenia, South Africa, Spain, Vietnam, USA and UK. There were 60 participants in total presenting 43 regular papers and 4 plenary presentations.

All regular participants at the workshop were invited to submit an eight page paper to be appear in the proceedings. Each paper was peer reviewed by two anonymous reviewers and only those accepted were invited to give presentations at the workshop.

The workshop was organised around topic based sessions consisting of fifteen minute presentations followed by questions and discussion. In addition, invited international plenary speakers gave special hour long keynote presentations on a range of topics based around the workshop themes. These were aimed to inspire new research directions both in theory and applications. To facilitate further debate and analysis an afternoon session was dedicated to a panel discussion of leading experts in the field and subsequent audience questions and comment.

There were four invited speakers covering topics ranging across the main themes of the workshop:
- Prof. Lotfi Zadeh, University of California at Berkley, USA
Title: Generalized Theory of Uncertainty - Principal Concepts and Ideas.
- Prof. Gert de Cooman, Ghent University, Belgium
Title: Reasoning with Vague Probability Assessments.
- Prof. Valdik Kreinovich, University of Texas at El Paso, USA
Title: Statistical Data Processing under Interval Uncertainty: Algorithms and Computational Complexity.
- Prof. Jim Hall, Tyndall Centre for Climate Change Research, and University of Newcastle upon Tyne, UK
Title: Soft Methods in Earth Systems Engineering.

The following sessions formed the core part of the workshop:
- Soft Methods in Statistics and Random Information Systems I (Chair: M. A. Gil, 6 papers)
-Soft Methods in Statistics and Random Information Systems II (Chair: P. Grzegorzewski, 5 papers)
- Applications and Modelling of Imprecise Operators (Special session organized and chaired by A. Bugarin from the University of Santiago de Compostela, Spain, 4 papers)
- Imprecise Probability Theory (Special session organized and chaired by E. Miranda from Rey Juan Carlos University, Spain, 6 paper)
- Probability of Imprecise-Valued Random Elements with Applications (Special session organized and chaired by S. Li from Beijing University of Technology, China, 5 papers)
- Possibility, Evidence and Interval Methods (Chair: J. Rossiter, 8 papers)
- Integrated Uncertainty Modelling in Applications (Chair J. Recasens, 9 papers)

A panel discussion focussed on the importance of the operational interpretation of membership functions in the future development of fuzzy set theory. The Panellists were: Prof. Yakov Ben-Haim (Technon-Israel Institute of Technology, Israel), Prof. Didier Dubois (Institut de Recherche de Informatique de Toulouse, France), Prof. Gert de Cooman (Ghent University, Belgium), Prof. Vladik Kreinovich (University of Texas, USA)
Exploitation Route The workshop was part of a series which has continued to run every two years making a significant contribution to research in this field.
Sectors Digital/Communication/Information Technologies (including Software),Environment,Healthcare

 
Description The uncertainty present in complex systems often has diverse sources and forms. It may be probabilistic in nature resulting from underlying random processes or may result from imprecision and fuzziness inherent in subjective human judgements and definitions. Hence, a realistic analysis providing decision makers with an accurate reflection of the true current state of knowledge requires an integrated framework incorporating both probability theory, statistics and soft methods. Formally representing imprecision and ignorance when they are indeed present provides less ad hoc and more robust numerical estimates of uncertainty measures. Such an integrated theory of uncertainty requires clear operational semantics for fundamental measures so that links between different methodologies can be identified. This requirement can motivate theoretical research with the potential to resolve a number of foundational issues concerning the interpretation and manipulation of uncertainty measures. SMPS 2006 provided a stimulating forum for discussion of both theoretical and practical issues of this kind. New methodologies and fundamental research issues where identified for all of the main themes.
First Year Of Impact 2006
Sector Digital/Communication/Information Technologies (including Software),Environment,Healthcare