Elicitation of expert opinions for uncertainty and risks—: Bilal M. Ayyub, CRC, Boca Raton, 2001, 302pp., ISBN 0-8493-1087-3
✍ Scribed by Roger M Cooke
- Publisher
- Elsevier Science
- Year
- 2003
- Tongue
- English
- Weight
- 71 KB
- Volume
- 133
- Category
- Article
- ISSN
- 0165-0114
No coin nor oath required. For personal study only.
✦ Synopsis
This book is divided into seven chapters. The ÿrst chapter "Knowledge and ignorance" is a review of theories of knowledge starting with the ÿrst Greek philosopher Thales (585 BC) and going up to the present, with many tables and block diagrams. The second chapter "Informationbased system deÿnition" deals with systems engineering in relation to knowledge, uncertainty and ignorance. Chapter 3 "Experts, opinions and elicitation methods" will be very familiar to readers of Experts in Uncertainty (Cooke, Oxford University Press, 1991). Chapter 4 "Expressing and modeling expert opinions" summarizes the many representations of uncertainty, including fuzzy sets, rough sets, evidence theory, probability and possibility theory. Chapter 5, entitled "Consensus and aggregating expert opinions" reviews the many ways of combining the many types of uncertainty identiÿed in Chapter 4. Chapter 6 contains "Guidance on expert-opinion elicitation", and the ÿnal chapter treats some applications. Each chapter contains exercises.
The book has a very wide sweep. The reader is presented a very wide range of possibilities for dealing with uncertainty and does not come away with a clear recommendation how to choose among them. Nor does the extended excursion into the theory of knowledge yield a framework for evaluating methods for dealing with uncertainty. While the book contains much useful encyclopedic material, there is, in my view, a signiÿcant selection bias at work. Thus, "interest in analyzing and modeling uncertainty and ignorance" is said to be "started by the works of Zadeh (1965 and 1978), Dempster (1976a, b), Shafer (1976), Sugeno (1974 and 1977), Klir and Folger (1988), Pawlak and Smithson (1989)." (p. 85). The reader is not told about the foundational work of Keynes, Borel, Von Mises, Ramsey, Von Neumann and Morgenstern, Popper, De Finetti, Savage, and many, many others. Modern philosophy of science is wholy absent. Semantic analysis, formulated by Mach and applied by Einstein and Bohr to enable the revolutions of relativity and quantum mechanics, has been a central theme in modern philosophy of science. Its absence is particularly unfortunate, as this broad tradition would supply ample conceptual tools for a critical evaluation of the many putative representations of uncertainty.
The technical body of the book is contained in Chapters 4 and 5. The exposition in Chapter 4 will not please the mathematicians. The "Fundamentals of classical set theory" (pp. 127-129) is very far from that. Thus, elements of sets are said to be either "discrete or continuous", which is meaningless without a topology (which one?). Sets are also said to be "convex or non-convex" which is meaningless without addition and scalar multiplication. The reader does not come away knowing what a set, or set theory, is. Much more attention is given to fuzzy sets and fuzzy arithmetic.
The exposition of fuzzy set theory leaves me totally confused. The fuzzy membership function A (x) : X → [0; 1] is said to represent the "degree of compatibility" of element x with set A.