************************************************************************** NEWSLETTER ON DECISION AND REASONING UNDER UNCERTAINTY Issue 99002 Editors: Salem Benferhat, Henri Prade 10.05.1999 Back issues available at http://www.ida.liu.se/ext/etai/dru/index.html ************************************************************************** This newsletter contains: 1. Creation of Directory of researchers 2. New book 3. Conference announcements ========================================================================== 1. Creation of Directory of researchers ========================================================================== A directory of about 400 people interested in DECISION AND REASONING UNDER UNCERTAINTY is temporarily located at: http://cafe.newcastle.edu.au/salem/chercheur.html Please send an email to benferhat@irit.fr or prade@irit.fr if there is any mistake. To subscribe, please send the following information: Last and first name, affiliation, email address, personal web-page Moreover, we encourage you to take advantage of this rather large special interest and discussion group and to submit papers to ETAI-DRU Journal. We recall that the call for papers is available at : http://www.ida.liu.se/ext/etai/dru/index.html Feel free to contact us for any further information. ========================================================================== 2. New book ========================================================================== Possibilistic Data Analysis for Operations Research Prof. Hideo Tanaka and Dr. Peijun Guo Studies in Fuzziness and Soft Computing Vol. 29, Physica-Verlag, 1999, 194pp, Hardcover, DM 98. http://www.springer.de/economics This monograph systematically introduces new theories and techniques of possibility theory for data analysis in operations research. As a counterpart of multivariate analysis based on probability theory, possibility data analysis plays an indispensable role in finance, economic, society fields and upper levels of decision-making system of industrial engineering. This book focuses on how to use possibility theory to analyze and model fuzzy or incomplete information inherently existing in real decision problems. The theoretical work in this book is original and practical, and problem-solving techniques are simple ones that are linear programming and quadratic programming. Readers not only can learn systematical, newest theories in possibility data analysis but also can learn the detailed methodologies to deal with the practical problems. The book can be used for anyone concerned with the methodologies and techniques on dealing with the uncertainty in real world problems, especially the people related with industrial engineering, informatics, finance engineering and economics and all people related with decision-making support system. The balance between the theoretical work and applications makes the book suitable for both researchers and engineers, as well as graduate students. Contents: Foreword by D. Dubois and H. Prade Chapter 1. Introduction : Possibility Theory in Operations Research 1.1. Possibility Distribution : A Knowledge Representation 1.2. The Role of Possibility Theory in Regression Analysis 1.3. The Role of Possibility Theory in Portfolio Selection Problems 1.4. The Role of Possibility Theory in Discriminant Analysis 1.5. The Roles of Possibility Theory in Other Topics 1.6. Chapter Description Chapter 2. Possibility Models 2.1. Possibility Distributions 2.2. Operations on Possibility Distributions 2.2.1. Interval Arithmetic 2.2.2. Fuzzy Number Arithmetic 2.2.3. Fuzzy Vector Arithmetic 2.3. Possibility and Necessity Measures 2.4. Probability Measures of Fuzzy Events 2.5. Possibilistic Linear Systems 2.6. Brief Bibliographical Remarks Chapter 3. Theory of Possibilistic Systems Based on Exponential Possibility Distributions 3.1. Combination Rule of Exponential Possibility Distributions 3.2. Marginal and Conditional Possibility Distributions 3.3. Continuous Fuzzy Relation Systems 3.4 Brief Bibliographical Remarks Chapter 4. Identification of Possibility Distributions 4.1. Principle of Maximum Likelihood 4.2. Identification of Upper and Lower Possibility Distributions 4.3. Numerical Examples 4.4. Brief Bibliographical Remarks Chapter 5. Possibilistic Regression Analysis 5.1. Statistical Regression analysis 5.2. Interval Regression 5.3. Fuzzy Regression 5.4. Exponential Possibilistic Regression 5.5. Interval Nonlinear Regression 5.6. Brief Bibliographical Remarks Chapter 6. Possibilistic Portfolio Selection Problems 6.1. Portfolio Selection Models Based on Probability Theory 6.1.1. Markowitz's Portfolio Selection Model 6.1.2. Models Based on Probability Measures 6.1.3 Model Based on Mean-Absolute Deviation 6.2. Portfolio Selection Models Based on Possibility Theory 6.2.1. Portfolio Selection Model Based on Aspiration Levels of Decision-Makers 6.2.2. Portfolio Selection Models Based on Possibility and Necessity Measures 6.3. Portfolio Selection Models Based on Fuzzy Probabilities 6.3.1. Definition of Fuzzy Probabilities 6.3.2. Fuzzy Probability Portfolio Selection Model 6.4. Portfolio Selection Models Based on Exponential Possibility Distributions 6.4.1. Identification of Possibility Distributions from Given Security Data 6.4.2. Portfolio Selection Model Based on Upper and Lower Possibility Distributions 6.4.3. Model Based on Necessity Measure 6.5. Numerical Examples 6.5.1. Portfolio Selection Based on Fuzzy Probabilities 6.5.2. Portfolio Selection Based on Upper and Lower Possibility Distributions 6.6. Brief Bibliographical Remarks Chapter 7. Discriminant Analysis Based on Possibility Distributions 7.1. discriminant Analysis by Bayes' Formula 7.2. Linear Discriminant Functions 7.3. Possibilistic discriminant Rules 7.4. Feature Vector for Classification by Possibility Measures 7.5. Possibilistic Classification for the Group of Data 7.6. Numerical Example 7.7. Brief Bibliographical Remarks Chapter 8. Rough Set Analysis 8.1. Basic Notions of Rough Sets 8.2. Reduction of Information Systems by Elementary Sets 8.3. Reduction of Information Systems by Accuracy Measures 8.4. Reduction for Divisions of Attributes 8.5. Fuzzy Inference Models 8.6. Fuzzy Expert System for Medical Diagnosis 8.7. Similarities between Rough Sets and Possibility Models 8.8. Brief Bibliographical Remarks ========================================================================== 3. Conference announcements ========================================================================== Workshop on Conditional Independence Structures and Graphical Models September 27 to October 1, 1999 The Fields Institute for Research in Mathematical Sciences, Toronto, Canada Organized within the Fields Institute Main Program Causal Interpretation and Identification of Conditional Independence Structures In collaboration with Institute of Information Theory and Automation Academy of Sciences of the Czech Republic SCIENTIFIC PROGRAMME 1. Philosophical and methodological foundations Development of the concept of conditional independence and related notions in statistics, philosophical logic, artificial intelligence, database theory and statistical physics. Historical aspects together with perspectives. 2. Conditional independence structures of graphical models Statistical and probabilistic models based on undirected graphs, acyclic directed graphs, chain graphs, reciprocal graphs, annotated graphs, alternative chain graphs, joint-response chain graphs, etc. Conditioning and marginalization. Data driven learning of the structure. Understanding latent variables. Bayesian aspects. 3. Information-theoretical approaches and non-graphical methods Formal properties of probabilistic conditional independence, information-theoretical inequalities and related problems of information theory and cryptology (secret-sharing). Among non-graphical models: lattice conditional independence models, theory of imsets, matroid-theoretical structures. 4. Conditional independence in other frameworks The concept of conditional independence in miscellaneous nonprobabilistic calculi; event trees, valuation networks, possibility theory and applications in artificial intelligence. Interval statistics, relational and probabilistic databases, coherence theory, decompositions based on conditional independence assumptions. MAIN SPEAKERS David Cox, Nuffield College, Oxford Phil Dawid, University College London, London Jan Koster, Erasmus University, Rotterdam Steffen Lauritzen, Aalborg University, Aalborg Azaria Paz, Technion Institute, Haifa Judea Pearl, University of California, Los Angeles Michael Perlman, University of Washington, Seattle Glenn Shafer, Rutgers University, New Jersey Prakash Shenoy, Kansas University Business School, Lawrence Nanny Wermuth, Center for Survey Research and Methodology, Mannheim Raymond Yeung, The Chinese University of Hong Kong, Hong Kong Zhen Zhang, University of Southern California, Los Angeles The programme will also contain contributed paper and poster sessions. The majority of lectures will be on invitation. No parallel sessions are planned. Contributions focusing on formulation of open mathematical problems will be prefered. SUBMISSION OF ABSTRACTS Participants willing to make a contribution should submit an abstract of a paper no longer than two pages in English or French. The deadline for the submission is June 30, 1999. Both ordinary and electronic submissions are accepted. If you are using the electronic submission, please download the LaTeX template here and restrain from using exotic LaTeX or TeX acrobatics. Alternatively, you can provide an ASCI file. With your abstract, please indicate which coauthor is going to present the paper, give the affiliation, email and conventional addresses of the authors and say whether the paper is intended for oral or poster presentation. All registered participants will be notified about acceptance or nonacceptance of their contributions before July 30th, 1999 and will obtain the book of abstracts upon arrival. Peer-reviewed proceedings containing full versions of contributed papers are likely to be edited after the workshop. IMPORTANT ADDRESSES Please, register by email if possible at cis@utia.cas.cz. Provide your full mailing address, email address and whether you want to obtain the second announcement, to participate and/or to present a contribution. Frantisek Matus & Milan Studeny Institute of Information Theory and Automation Academy of Sciences of the Czech Republic Pod vodarenskou vezi 4 18208 Prague, Czech Republic Phone: (420-2) 6605 2341 Fax: (420-2) 688 4903 E-mail: cis@utia.cas.cz IMPORTANT DATES June 30, 1999 Deadline for Abstracts July 30, 1999 Notification of Acceptance September 27, 1999 Workshop Opening Day More details can be found: http://www.utia.cas.cz/user_data/matus/toronto'99/workshop.html --------------------------------------- 2 . The 5th Joint Conference on Information Sciences (JCIS 2000) to be held in Atlantic City, New Jersey, February 27-March 3, 2000, actually is the 9th conference sponsored by Duke University and the Association for Intelligent Machinery, Inc. (AIM). The first four conferences dealt primarily with fuzzy technology and neural networks. These were followed by joint conferences on all aspects of information science and intelligent systems. The acceptance of this joint effort by many researchers all over the world says much about its appeal and the need for such an important gathering. Not only does this "one-stop shopping" save valuable resources for attendees, there are some researchers whose primary emphasis incorporates several methodologies simultaneously. Our goal can be summarized in two words-integration and commercialization. Cognitive science is an historical example of this type of integration. We see our role as supplementary to the efforts of major establishments that provide invaluable service to our research community. We are serving a small percentage of researchers working with emerging technologies that integrate system components for successful and marketable intelligent systems. Selected conference papers-grouped as special issues-will be published in Information Sciences Journal, which has the similar goal of integration and technology transfer. We also are working diligently to negotiate an agreement with a major publisher for an intelligent systems series containing peer-reviewed papers received at the JCIS 2000. Please visit our web site for the latest conference information: www.ee.duke.edu/JCIS/. This information is being updated frequently as our conference organizers are inviting more leaders in related fields, including Noble Laureates. To date, the following leaders in our scientific community have accepted invitations to deliver keynote or plenary speeches. Jim Anderson T. S. Huang Karl H. Pribram Wolfgang Banzhaf Janusz Kacprzyk Jeffrey P. Sutton B. Chandrasekaran A. C. Kak Ron Yager Lawrence J. Fogel S. C. Kak Walter J. Freeman John Mordeson David E. Goldberg Kumpati S. Narenda Irwin Goodman Anil Nerode Stephen Grossberg Huang T. Nguyen Azriel Rosenfeld and Lotfi Zadeh are honorary conference chairs. JCIS 2000 includes nine separate tracks. Contacts are listed next to each conference or workshop: 7th International Conference on Fuzzy Theory and Technology Co-Chairs: Janusz Kacprzyk, Ron R. Yager Program Co-Chairs: Dan Schwartz (schwartz@nu.cs.fsu.edu), Sujeet Shenoi 5th International Conference on Computer Science and Informatics Co-Chairs: Mi Lu, Keqin Li Program Chair: Mi Lu (mlu@ee.tamu.edu) 4th International Conference on Computational Intelligence and Neurosciences Co-Chairs: Jeffrey Sutton, Subhash C. Kak Program Co-Chairs: George Georgiou (georgiou@csci.csusb.edu), Bruce MacLellan 3rd International Workshop on Intelligent Control and Systems Chair: Kumpati Narendra Program Co-Chairs: Hao Ying (hying@utmb.edu), Paul P. Wang 3rd International Workshop on Frontiers in Evolutionary Algorithms Chair: David Goldberg Program Chair: Manuel Grana Romay (ccpgrrom@si.ehu.es) 3rd International Conference on Computer Vision, Pattern Recognition and Image Processing Chair and Program Chair: Heng-Da Cheng (cheng@hengda.cs.usu.edu) 1st International Workshop on Intelligent Multimedia Computing and Networking Chair and Program Chair: Timothy K. Shih (tshih@cs.tku.edu.tw) 2nd International Workshop on Biomolecular Informatics Co-Chairs and Program Co-Chairs: James R. Gattiker, Jason T. L. Wang (jason@cis.njit.edu) 1st International Workshop on Intelligent Systems for Speech and Language: Speech and Language Acquisition, Recognition, and Understanding in Brains and Machines Chair: Nikola Kasabov (nkasabov@otago.ac.nz) Important Dates September 1, 1999 - Deadline for submission of paper summaries. November 1, 1999 - Paper acceptance letters out to authors upon the review of summaries December 1, 1999 - Deadline for early registration with discounted fee. December 1, 1999 - Deposit due for each paper to be included in proceedings December 1, 1999 - Deadline for submission of revised summaries, camera ready for publication December 1, 1999 - Deadline for invited sessions and exhibition proposals February 27, 2000 - Opening of JCIS 2000 Please visit our web site for conference fees and hotel information. If you have further questions, you may contact us via e-mail at jcis@ee.duke.edu or write to us at JCIS 2000, Department of Electrical and Computer Engineering, Duke University, Box 90291, Durham, NC 27708 USA. We look forward to seeing you there! Paul P. Wang General Chair JCIS 2000