Search

This book serves as an advanced text for a graduate course on stochastic algorithms for the students of probability and statistics, engineering, economics and machine learning. This second edition gives a comprehensive treatment of stochastic approximation algorithms based on the ordinary differential equation (ODE) approach which analyses the algorithm in terms of a limiting ODE. It has a streamlined treatment of the classical convergence analysis and includes several recent developments such as concentration bounds, avoidance of traps, stability tests, distributed and asynchronous schemes, multiple time scales, general noise models, etc., and a category-wise exposition of many important applications. It is also a useful reference for researchers and practitioners in the field.

An Introduction to Universal Artificial Intelligence provides the formal underpinning of what it means for an agent to act intelligently in an unknown environment. First presented in Universal Algorithmic Intelligence (Hutter, 2000), UAI offers a framework in which virtually all AI problems can be formulated, and a theory of how to solve them. UAI unifies ideas from sequential decision theory, Bayesian inference, and algorithmic information theory to construct AIXI, an optimal reinforcement learning agent that learns to act optimally in unknown environments. AIXI is the theoretical gold standard for intelligent behavior.The book covers both the theoretical and practical aspects of UAI. Bayesian updating can be done efficiently with context tree weighting, and planning can be approximated by sampling with Monte Carlo tree search. It provides algorithms for the reader to implement, and experimental results to compare against. These algorithms are used to approximate AIXI. The book ends with a philosophical discussion of Artificial General Intelligence: Can super-intelligent agents even be constructed? Is it inevitable that they will be constructed, and what are the potential consequences?This text is suitable for late undergraduate students. It provides an extensive chapter to fill in the required mathematics, probability, information, and computability theory background.

This book constitutes the refereed proceedings of the 8th International Conference on Belief Functions, BELIEF 2024, held in Belfast, UK, in September 2–4, 2024. The 30 full papers presented in this book were carefully selected and reviewed from 36 submissions. The papers cover a wide range on theoretical aspects on Machine learning; Statistical inference; Information fusion and optimization; Measures of uncertainty, conflict and distances; Continuous belief functions, logics, computation.

This book serves as an advanced text for a graduate course on stochastic algorithms for graduate students in probability and statistics, engineering, economics and machine learning. This second edition gives a comprehensive treatment of stochastic approximation algorithms based on the “ordinary differential equation (ODE) approach” which analyses the algorithm in terms of a limiting ODE. It has a streamlined treatment of the classical convergence analysis and includes several recent developments such as concentration bounds, avoidance of traps, stability tests, distributed and asynchronous schemes, multiple time scales, general noise models, etc., and a category-wise exposition of many important applications. It is also a useful reference for researchers and practitioners in the field.

Professor Judea Pearl won the 2011 Turing Award “for fundamental contributions to artificial intelligence through the development of a calculus for probabilistic and causal reasoning.” This book contains the original articles that led to the award, as well as other seminal works, divided into four parts: heuristic search, probabilistic reasoning, causality, first period (1988–2001), and causality, recent period (2002–2020). Each of these parts starts with an introduction written by Judea Pearl. The volume also contains original, contributed articles by leading researchers that analyze, extend, or assess the influence of Pearl's work in different fields: from AI, Machine Learning, and Statistics to Cognitive Science, Philosophy, and the Social Sciences. The first part of the volume includes a biography, a transcript of his Turing Award Lecture, two interviews, and a selected bibliography annotated by him.

Recent advances in the area of lifted inference, which exploits the structure inherent in relational probabilistic models.Statistical relational AI (StaRAI) studies the integration of reasoning under uncertainty with reasoning about individuals and relations. The representations used are often called relational probabilistic models. Lifted inference is about how to exploit the structure inherent in relational probabilistic models, either in the way they are expressed or by extracting structure from observations. This book covers recent significant advances in the area of lifted inference, providing a unifying introduction to this very active field.After providing necessary background on probabilistic graphical models, relational probabilistic models, and learning inside these models, the book turns to lifted inference, first covering exact inference and then approximate inference. In addition, the book considers the theory of liftability and acting in relational domains, which allows the connection of learning and reasoning in relational domains.ContributorsBabak Ahmadi, Hendrik Blockeel, Hung Bui, Yuqiao Chen, Arthur Choi, Jaesik Choi, Adnan Darwiche, Jesse Davis, Rodrigo de Salvo Braz, Pedro Domingos, Daan Fierens, Martin Grohe, Fabian Hadiji, Seyed Mehran Kazemi, Kristian Kersting, Roni Khardon, Angelika Kimmig, Jacek Kisynski, Daniel Lowd, Wannes Meert, Martin Mladenov, Raymond Mooney, Sriraam Natarajan, Mathias Niepert, David Poole, Scott Sanner, Pascal Schweitzer, Nima Taghipour, Guy Van den Broeck

This book introduces systematically the application of Bayesian probabilistic approach in soil mechanics and geotechnical engineering. Four typical problems are analyzed by using Bayesian probabilistic approach, i.e., to model the effect of initial void ratio on the soil–water characteristic curve (SWCC) of unsaturated soil, to select the optimal model for the prediction of the creep behavior of soft soil under one-dimensional straining, to identify model parameters of soils and to select constitutive model of soils considering critical state concept. This book selects the simple and easy-to-understand Bayesian probabilistic algorithm, so that readers can master the Bayesian method to analyze and solve the problem in a short time. In addition, this book provides MATLAB codes for various algorithms and source codes for constitutive models so that readers can directly analyze and practice.This book is useful as a postgraduate textbook for civil engineering, hydraulic engineering, transportation, railway, engineering geology and other majors in colleges and universities, and as an elective course for senior undergraduates. It is also useful as a reference for relevant professional scientific researchers and engineers.

The principal aim of this book is to introduce to the widest possible audience an original view of belief calculus and uncertainty theory. In this geometric approach to uncertainty, uncertainty measures can be seen as points of a suitably complex geometric space, and manipulated in that space, for example, combined or conditioned. In the chapters in Part I, Theories of Uncertainty, the author offers an extensive recapitulation of the state of the art in the mathematics of uncertainty. This part of the book contains the most comprehensive summary to date of the whole of belief theory, with Chap. 4 outlining for the first time, and in a logical order, all the steps of the reasoning chain associated with modelling uncertainty using belief functions, in an attempt to provide a self-contained manual for the working scientist. In addition, the book proposes in Chap. 5 what is possibly the most detailed compendium available of all theories of uncertainty. Part II, The Geometry of Uncertainty, is the core of this book, as it introduces the author's own geometric approach to uncertainty theory, starting with the geometry of belief functions: Chap. 7 studies the geometry of the space of belief functions, or belief space, both in terms of a simplex and in terms of its recursive bundle structure; Chap. 8 extends the analysis to Dempster's rule of combination, introducing the notion of a conditional subspace and outlining a simple geometric construction for Dempster's sum; Chap. 9 delves into the combinatorial properties of plausibility and commonality functions, as equivalent representations of the evidence carried by a belief function; then Chap. 10 starts extending the applicability of the geometric approach to other uncertainty measures, focusing in particular on possibility measures (consonant belief functions) and the related notion of a consistent belief function. The chapters in Part III, Geometric Interplays, are concerned with the interplay of uncertainty measures of different kinds, and the geometry of their relationship, with a particular focus on the approximation problem. Part IV, Geometric Reasoning, examines the application of the geometric approach to the various elements of the reasoning chain illustrated in Chap. 4, in particular conditioning and decision making. Part V concludes the book by outlining a future, complete statistical theory of random sets, future extensions of the geometric approach, and identifying high-impact applications to climate change, machine learning and artificial intelligence. The book is suitable for researchers in artificial intelligence, statistics, and applied science engaged with theories of uncertainty. The book is supported with the most comprehensive bibliography on belief and uncertainty theory.

This book presents selected peer-reviewed contributions from the International Conference on Time Series and Forecasting, ITISE 2018, held in Granada, Spain, on September 19-21, 2018. The first three parts of the book focus on the theory of time series analysis and forecasting, and discuss statistical methods, modern computational intelligence methodologies, econometric models, financial forecasting, and risk analysis. In turn, the last three parts are dedicated to applied topics and include papers on time series analysis in the earth sciences, energy time series forecasting, and time series analysis and prediction in other real-world problems. The book offers readers valuable insights into the different aspects of time series analysis and forecasting, allowing them to benefit both from its sophisticated and powerful theory, and from its practical applications, which address real-world problems in a range of disciplines. The ITISE conference series provides avaluable forum for scientists, engineers, educators and students to discuss the latest advances and implementations in the field of time series analysis and forecasting. It focuses on interdisciplinary and multidisciplinary research encompassing computer science, mathematics, statistics and econometrics.

This book comprises nine selected works on numerical and computational methods for solving multiobjective optimization, game theory, and machine learning problems. It provides extended versions of selected papers from various fields of science such as computer science, mathematics and engineering that were presented at EVOLVE 2013 held in July 2013 at Leiden University in the Netherlands. The internationally peer-reviewed papers include original work on important topics in both theory and applications, such as the role of diversity in optimization, statistical approaches to combinatorial optimization, computational game theory, and cell mapping techniques for numerical landscape exploration. Applications focus on aspects including robustness, handling multiple objectives, and complex search spaces in engineering design and computational biology.

The aim of this book is to provide an introduction to probability logic-based formalization of uncertain reasoning. The authors'primary interest is mathematical techniques for infinitary probability logics used to obtain results about proof-theoretical and model-theoretical issues such as axiomatizations, completeness, compactness, and decidability, including solutions of some problems from the literature. An extensive bibliography is provided to point to related work, and this book may serve as a basis for further research projects, as a reference for researchers using probability logic, and also as a textbook for graduate courses in logic.

Comprehensively covers protein subcellular localization from single-label prediction to multi-label prediction, and includes prediction strategies for virus, plant, and eukaryote species. Three machine learning tools are introduced to improve classification refinement, feature extraction, and dimensionality reduction.

This accessible text/reference provides a general introduction to probabilistic graphical models (PGMs) from an engineering perspective. The book covers the fundamentals for each of the main classes of PGMs, including representation, inference and learning principles, and reviews real-world applications for each type of model. These applications are drawn from a broad range of disciplines, highlighting the many uses of Bayesian classifiers, hidden Markov models, Bayesian networks, dynamic and temporal Bayesian networks, Markov random fields, influence diagrams, and Markov decision processes. Features: presents a unified framework encompassing all of the main classes of PGMs; describes the practical application of the different techniques; examines the latest developments in the field, covering multidimensional Bayesian classifiers, relational graphical models and causal models; provides exercises, suggestions for further reading, and ideas for research or programming projects at the end of each chapter.

In recent years it has become apparent that an important part of the theory of Artificial Intelligence is concerned with reasoning on the basis of uncertain, incomplete or inconsistent information. Classical logic and probability theory are only partially adequate for this, and a variety of other formalisms have been developed, some of the most important being fuzzy methods, possibility theory, belief function theory, non monotonic logics and modal logics. The aim of this workshop was to contribute to the elucidation of similarities and differences between the formalisms mentioned above.

This book focuses on the state of the art of Monte Carlo methods in radiation physics and particle transport simulation and applications. Special attention is paid to algorithm development for modeling, and the analysis of experiments and measurements in a variety of fields.

This volume, the 6th volume in the DRUMS Handbook series, is part of the after­ math of the successful ESPRIT project DRUMS (Defeasible Reasoning and Un­ certainty Management Systems) which took place in two stages from 1989-1996. In the second stage (1993-1996) a work package was introduced devoted to the topics Reasoning and Dynamics, covering both the topics of'Dynamics of Rea­ soning', where reasoning is viewed as a process, and'Reasoning about Dynamics', which must be understood as pertaining to how both designers of and agents within dynamic systems may reason about these systems. The present volume presents work done in this context. This work has an emphasis on modelling and formal techniques in the investigation of the topic'Reasoning and Dynamics', but it is not mere theory that occupied us. Rather research was aimed at bridging the gap between theory and practice. Therefore also real-life applications of the modelling techniques were considered, and we hope this also shows in this volume, which is focused on the dynamics of reasoning processes. In order to give the book a broader perspective, we have invited a number of well-known researchers outside the project but working on similar topics to contribute as well. We have very pleasant recollections of the project, with its lively workshops and other meetings, with the many sites and researchers involved, both within and outside our own work package.

In both the linguistic and the language engineering community, the creation and use of annotated text collections (or annotated corpora) is currently a hot topic. Annotated texts are of interest for research as well as for the development of natural language pro­ cessing (NLP) applications. Unfortunately, the annotation of text material, especially more interesting linguistic annotation, is as yet a difficult task and can entail a substan­ tial amount of human involvement. Allover the world, work is being done to replace as much as possible of this human effort by computer processing. At the frontier of what can already be done (mostly) automatically we find syntactic wordclass tagging, the annotation of the individual words in a text with an indication of their morpho syntactic classification. This book describes the state of the art in syntactic wordclass tagging. As an attempt to give an overall view of the field, this book is of interest to (at least) two, possibly very different, types of reader. The first type consists of those people who are using, or are planning to use, tagged material and taggers. They will want to know what the possibilities and impossibilities of tagging are, but are not necessarily interested in the internal working of automatic taggers. This, on the other hand, is the main interest of our second type of reader, the builders of automatic taggers and other natural language processing software.

Foundations of Bayesianism is an authoritative collection of papers addressing the key challenges that face the Bayesian interpretation of probability today. Some of these papers seek to clarify the relationships between Bayesian, causal and logical reasoning. Others consider the application of Bayesianism to artificial intelligence, decision theory, statistics and the philosophy of science and mathematics. The volume includes important criticisms of Bayesian reasoning and also gives an insight into some of the points of disagreement amongst advocates of the Bayesian approach. The upshot is a plethora of new problems and directions for Bayesians to pursue. The book will be of interest to graduate students or researchers who wish to learn more about Bayesianism than can be provided by introductory textbooks to the subject. Those involved with the applications of Bayesian reasoning will find essential discussion on the validity of Bayesianism and its limits, while philosophers and others interested in pure reasoning will find new ideas on normativity and the logic of belief.

Intelligent Information Technology (iiT) encompasses the theories and ap­ plications of artificial intelligence, statistical pattern recognition, learning theory, data warehousing, data mining and knowledge discovery, Grid com­ puting, and autonomous agents and multi-agent systems in the context of today's as well as future IT, such as Electronic Commerce (EC), Business Intelligence (BI), Social Intelligence (SI), Web Intelligence (WI), Knowledge Grid (KG), and Knowledge Community (KC), among others. The multi-author monograph presents the current state of the research and development in intelligent technologies for information analysis, in par­ ticular, advances in agents, data mining, and learning theory, from both the­ oretical and application aspects. It investigates the future of information technology (IT) from a new intelligent IT (iiT) perspective, and highlights major iiT-related topics by structuring an introductory chapter and 22 sur­ vey/research chapters into 5 parts: (1) emerging data mining technology, (2) data mining for Web intelligence, (3) emerging agent technology, ( 4) emerging soft computing technology, and (5) statistical learning theory. Each chapter includes the original work of the author(s) as well as a comprehensive survey related to the chapter's topic. This book will become a valuable source of reference for R&D profession­ als active in advanced intelligent information technologies. Students as well as IT professionals and ambitious practitioners concerned with advanced in­ telligent information technologies will appreciate the book as a useful text enhanced by numerous illustrations and examples.

Classical probability theory and mathematical statistics appear sometimes too rigid for real life problems, especially while dealing with vague data or imprecise requirements. These problems have motivated many researchers to'soften'the classical theory. Some'softening'approaches utilize concepts and techniques developed in theories such as fuzzy sets theory, rough sets, possibility theory, theory of belief functions and imprecise probabilities, etc. Since interesting mathematical models and methods have been proposed in the frameworks of various theories, this text brings together experts representing different approaches used in soft probability, statistics and data analysis.

This book contains leading survey papers on the various aspects of Abduction, both logical and numerical approaches. Abduction is central to all areas of applied reasoning, including artificial intelligence, philosophy of science, machine learning, data mining and decision theory, as well as logic itself.

The last decade has seen a remarkable development of the'Marginal and Moment Problems'as a research area in Probability and Statistics. Its attractiveness stemmed from its lasting ability to provide a researcher with difficult theoretical problems that have direct consequences for appli­ cations outside of mathematics. The relevant research aims centered mainly along the following lines that very frequently met each other to provide sur­ prizing and useful results : -To construct a probability distribution (to prove its existence, at least) with a given support and with some additional inner stochastic property defined typically either by moments or by marginal distributions. -To study the geometrical and topological structure of the set of prob­ ability distributions generated by such a property mostly with the aim to propose a procedure that would result in a stochastic model with some optimal properties within the set of probability distributions. These research aims characterize also, though only very generally, the scientific program of the 1996 conference'Distributions with given marginals and moment problems'held at the beginning of September in Prague, Czech Republic, to perpetuate the tradition and achievements of the closely related 1990 Roma symposium'On Frechet Classes'1 and 1993 Seattle'AMS Summer Conference on Marginal Problem'.

Searching is an important process in most AI systems, especially in those AI production systems consisting of a global database, a set of production rules, and a control system. Because of the intractability of uninformed search procedures, the use of heuristic information is necessary in most searching processes of AI systems. This important concept of heuristic informatioD is the central topic of this book. We first use the 8-puzzle and the game tic-tac-toe (noughts and crosses) as examples to help our discussion. The 8-puzzle consists of eight numbered movable tiles set in a 3 x 3 frame. One cell of the frame is empty so that it is possible to move an adjacent numbered tile into the empty cell. Given two tile configurations, initial and goal, an 8-puzzle problem consists of changing the initial configuration into the goal configuration, as illustrated in Fig. 1.1. A solution to this problem is a sequence of moves leading from the initial configuration to the goal configuration, and an optimal solution is a solution having the smallest number of moves. Not all problems have solutions; for example, in Fig. 1.1, Problem 1 has many solutions while Problem 2 has no solution at all.

This volume has its origin in the Seventeenth International Workshop on Maximum Entropy and Bayesian Methods, MAXENT 97. The workshop was held at Boise State University in Boise, Idaho, on August 4 -8, 1997. As in the past, the purpose of the workshop was to bring together researchers in different fields to present papers on applications of Bayesian methods (these include maximum entropy) in science, engineering, medicine, economics, and many other disciplines. Thanks to significant theoretical advances and the personal computer, much progress has been made since our first Workshop in 1981. As indicated by several papers in these proceedings, the subject has matured to a stage in which computational algorithms are the objects of interest, the thrust being on feasibility, efficiency and innovation. Though applications are proliferating at a staggering rate, some in areas that hardly existed a decade ago, it is pleasing that due attention is still being paid to foundations of the subject. The following list of descriptors, applicable to papers in this volume, gives a sense of its contents: deconvolution, inverse problems, instrument (point-spread) function, model comparison, multi sensor data fusion, image processing, tomography, reconstruction, deformable models, pattern recognition, classification and group analysis, segmentation/edge detection, brain shape, marginalization, algorithms, complexity, Ockham's razor as an inference tool, foundations of probability theory, symmetry, history of probability theory and computability. MAXENT 97 and these proceedings could not have been brought to final form without the support and help of a number of people.

Image analysis is one of the most challenging areas in today's computer sci­ ence, and image technologies are used in a host of applications. This book concentrates on image textures and presents novel techniques for their sim­ ulation, retrieval, and segmentation using specific Gibbs random fields with multiple pairwise interaction between signals as probabilistic image models. These models and techniques were developed mainly during the previous five years (in relation to April 1999 when these words were written). While scanning these pages you may notice that, in spite of long equa­ tions, the mathematical background is extremely simple. I have tried to avoid complex abstract constructions and give explicit physical (to be spe­ cific,'image-based') explanations to all the mathematical notions involved. Therefore it is hoped that the book can be easily read both by professionals and graduate students in computer science and electrical engineering who take an interest in image analysis and synthesis. Perhaps, mathematicians studying applications of random fields may find here some less traditional, and thus controversial, views and techniques.

Inspired by the eternal beauty and truth of the laws governing the run of stars on heavens over his head, and spurred by the idea to catch, perhaps for the smallest fraction of the shortest instant, the Eternity itself, man created such masterpieces of human intellect like the Platon's world of ideas manifesting eternal truths, like the Euclidean geometry, or like the Newtonian celestial me­ chanics. However, turning his look to the sub-lunar world of our everyday efforts, troubles, sorrows and, from time to time but very, very seldom, also our successes, he saw nothing else than a world full of uncertainty and tem­ porariness. One remedy or rather consolation was that of the deep and sage resignation offered by Socrates: I know, that I know nothing. But, happy or unhappy enough, the temptation to see and to touch at least a very small por­ tion of eternal truth also under these circumstances and behind phenomena charged by uncertainty was too strong. Probability theory in its most sim­ ple elementary setting entered the scene. It happened in the same, 17th and 18th centuries, when celestial mechanics with its classical Platonist paradigma achieved its greatest triumphs. The origins of probability theory were inspired by games of chance like roulettes, lotteries, dices, urn schemata, etc. and probability values were simply defined by the ratio of successful or winning results relative to the total number of possible outcomes.

Information usually comes in pieces, from different sources. It refers to different, but related questions. Therefore information needs to be aggregated and focused onto the relevant questions. Considering combination and focusing of information as the relevant operations leads to a generic algebraic structure for information. This book introduces and studies information from this algebraic point of view. Algebras of information provide the necessary abstract framework for generic inference procedures. They allow the application of these procedures to a large variety of different formalisms for representing information. At the same time they permit a generic study of conditional independence, a property considered as fundamental for knowledge presentation. Information algebras provide a natural framework to define and study uncertain information. Uncertain information is represented by random variables that naturally form information algebras. This theory also relates to probabilistic assumption-based reasoning in information systems and is the basis for the belief functions in the Dempster-Shafer theory of evidence.

Aggregation plays a central role in many of the technological tasks we are faced with. The importance of this process will become even greater as we move more and more toward becoming an information-cent.ered society, us is happening with the rapid growth of the Internet and the World Wirle Weh. Here we shall be faced with many issues related to the fusion of information. One very pressing issue here is the development of mechanisms to help search for information, a problem that clearly has a strong aggregation-related component. More generally, in order to model the sophisticated ways in which human beings process information, as well as going beyond the human capa­ bilities, we need provide a basket of aggregation tools. The centrality of aggregation in human thought can be be very clearly seen by looking at neural networks, a technology motivated by modeling the human brain. One can see that the basic operations involved in these networks are learning and aggregation. The Ordered Weighted Averaging (OWA) operators provide a parameter­ ized family of aggregation operators which include many of the well-known operators such as the maximum, minimum and the simple average.

Artificial neural networks and genetic algorithms both are areas of research which have their origins in mathematical models constructed in order to gain understanding of important natural processes. By focussing on the process models rather than the processes themselves, significant new computational techniques have evolved which have found application in a large number of diverse fields. This diversity is reflected in the topics which are the subjects of contributions to this volume. There are contributions reporting theoretical developments in the design of neural networks, and in the management of their learning. In a number of contributions, applications to speech recognition tasks, control of industrial processes as well as to credit scoring, and so on, are reflected. Regarding genetic algorithms, several methodological papers consider how genetic algorithms can be improved using an experimental approach, as well as by hybridizing with other useful techniques such as tabu search. The closely related area of classifier systems also receives a significant amount of coverage, aiming at better ways for their implementation. Further, while there are many contributions which explore ways in which genetic algorithms can be applied to real problems, nearly all involve some understanding of the context in order to apply the genetic algorithm paradigm more successfully. That this can indeed be done is evidenced by the range of applications covered in this volume.

In contrast to the prevailing tradition in epistemology, the focus in this book is on low-level inferences, i.e., those inferences that we are usually not consciously aware of and that we share with the cat nearby which infers that the bird which she sees picking grains from the dirt, is able to fly. Presumably, such inferences are not generated by explicit logical reasoning, but logical methods can be used to describe and analyze such inferences. Part 1 gives a purely system-theoretic explication of belief and inference. Part 2 adds a reliabilist theory of justification for inference, with a qualitative notion of reliability being employed. Part 3 recalls and extends various systems of deductive and nonmonotonic logic and thereby explains the semantics of absolute and high reliability. In Part 4 it is proven that qualitative neural networks are able to draw justified deductive and nonmonotonic inferences on the basis of distributed representations. This is derived from a soundness/completeness theorem with regard to cognitive semantics of nonmonotonic reasoning. The appendix extends the theory both logically and ontologically, and relates it to A. Goldman's reliability account of justified belief.

Belief change is an emerging field of artificial intelligence and information science dedicated to the dynamics of information and the present book provides a state-of-the-art picture of its formal foundations. It deals with the addition, deletion and combination of pieces of information and, more generally, with the revision, updating and fusion of knowledge bases. The book offers an extensive coverage of, and seeks to reconcile, two traditions in the kinematics of belief that often ignore each other - the symbolic and the numerical (often probabilistic) approaches. Moreover, the work encompasses both revision and fusion problems, even though these two are also commonly investigated by different communities. Finally, the book presents the numerical view of belief change, beyond the probabilistic framework, covering such approaches as possibility theory, belief functions and convex gambles. The work thus presents a unified view of belief change operators, drawing from a widely scattered literature embracing philosophical logic, artificial intelligence, uncertainty modelling and database systems. The material is a clearly organised guide to the literature on the dynamics of epistemic states, knowledge bases and uncertain information, suitable for scholars and graduate students familiar with applied logic, knowledge representation and uncertain reasoning.

In 1978 Edwin T. Jaynes and Myron Tribus initiated a series of workshops to exchange ideas and recent developments in technical aspects and applications of Bayesian probability theory. The first workshop was held at the University of Wyoming in 1981 organized by C.R. Smith and W.T. Grandy. Due to its success, the workshop was held annually during the last 18 years. Over the years, the emphasis of the workshop shifted gradually from fundamental concepts of Bayesian probability theory to increasingly realistic and challenging applications. The 18th international workshop on Maximum Entropy and Bayesian Methods was held in Garching / Munich (Germany) (27-31. July 1998). Opening lectures by G. Larry Bretthorst and by Myron Tribus were dedicated to one of th the pioneers of Bayesian probability theory who died on the 30 of April 1998: Edwin Thompson Jaynes. Jaynes revealed and advocated the correct meaning of'probability'as the state of knowledge rather than a physical property. This inter­ pretation allowed him to unravel longstanding mysteries and paradoxes. Bayesian probability theory,'the logic of science'- as E.T. Jaynes called it - provides the framework to make the best possible scientific inference given all available exper­ imental and theoretical information. We gratefully acknowledge the efforts of Tribus and Bretthorst in commemorating the outstanding contributions of E.T. Jaynes to the development of probability theory.

Decision Science and Technology is a compilation of chapters written in honor of a remarkable man, Ward Edwards. Among Ward's many contributions are two significant accomplishments, either of which would have been enough for a very distinguished career. First, Ward is the founder of behavioral decision theory. This interdisciplinary discipline addresses the question of how people actually confront decisions, as opposed to the question of how they should make decisions. Second, Ward laid the groundwork for sound normative systems by noticing which tasks humans can do well and which tasks computers should perform. This volume, organized into five parts, reflects those accomplishments and more. The book is divided into four sections: `Behavioral Decision Theory'examines theoretical descriptions and empirical findings about human decision making. `Decision Analysis'examines topics in decision analysis.`Decision in Society'explores issues in societal decision making. The final section, `Historical Notes', provides some historical perspectives on the development of the decision theory. Within these sections, major, multi-disciplinary scholars in decision theory have written chapters exploring some very bold themes in the field, as an examination of the book's contents will show. The main reason for the health of the Decision Analysis field is its close links between theory and applications that have characterized it over the years. In this volume, the chapters by Barron and Barrett; Fishburn; Fryback; Keeney; Moreno, Pericchi, and Kadane; Howard; Phillips; Slovic and Gregory; Winkler; and, above all, von Winterfeldt focus on those links. Decision science originally developed out of concern with real decision problems; and applied work, such as is represented in this volume, will help the field to remain strong.

The approach to probability theory followed in this book (which differs radically from the usual one, based on a measure-theoretic framework) characterizes probability as a linear operator rather than as a measure, and is based on the concept of coherence, which can be framed in the most general view of conditional probability. It is a `flexible'and unifying tool suited for handling, e.g., partial probability assessments (not requiring that the set of all possible `outcomes'be endowed with a previously given algebraic structure, such as a Boolean algebra), and conditional independence, in a way that avoids all the inconsistencies related to logical dependence (so that a theory referring to graphical models more general than those usually considered in bayesian networks can be derived). Moreover, it is possible to encompass other approaches to uncertain reasoning, such as fuzziness, possibility functions, and default reasoning. The book is kept self-contained, provided the reader is familiar with the elementary aspects of propositional calculus, linear algebra, and analysis.

'Statistical Modeling, Analysis and Management of Fuzzy Data,'or SMFD for short, is an important contribution to a better understanding of a basic issue -an issue which has been controversial, and still is though to a lesser degree. In substance, the issue is: are fuzziness and randomness distinct or coextensive facets of uncertainty? Are the theories of fuzziness and random­ ness competitive or complementary? In SMFD, these and related issues are addressed with rigor, authority and insight by prominent contributors drawn, in the main, from probability theory, fuzzy set theory and data analysis com­ munities. First, a historical perspective. The almost simultaneous births -close to half a century ago-of statistically-based information theory and cybernetics were two major events which marked the beginning of the steep ascent of probability theory and statistics in visibility, influence and importance. I was a student when information theory and cybernetics were born, and what is etched in my memory are the fascinating lectures by Shannon and Wiener in which they sketched their visions of the coming era of machine intelligence and automation of reasoning and decision processes. What I heard in those lectures inspired one of my first papers (1950)'An Extension of Wiener's Theory of Prediction,'and led to my life-long interest in probability theory and its applications to information processing, decision analysis and control.

This monograph is the r st in Fuzzy Approximation Theory. It contains mostly the author s research work on fuzziness of the last ten years and relies a lot on [10]-[32] and it is a natural outgrowth of them. It belongs to the broader area of Fuzzy Mathematics. Chapters are self-contained and several advanced courses can be taught out of this book. We provide lots of applications but always within the framework of Fuzzy Mathematics. In each chapter is given background and motivations. A c- plete list of references is provided at the end. The topics covered are very diverse. In Chapter 1 we give an extensive basic background on Fuzziness and Fuzzy Real Analysis, as well a complete description of the book. In the following Chapters 2,3 we cover in deep Fuzzy Di?erentiation and Integ- tion Theory, e.g. we present Fuzzy Taylor Formulae. It follows Chapter 4 on Fuzzy Ostrowski Inequalities. Then in Chapters 5, 6 we present results on classical algebraic and trigonometric polynomial Fuzzy Approximation.

The idea that belief comes in degrees is based on the observation that we are more certain of some things than of others. Various theories try to give accounts of how measures of this confidence do or ought to behave, both as far as the internal mental consistency of the agent as well as his betting, or other, behaviour is concerned. This anthology is the first book to give a balanced overview of these theories. It also explicitly relates these debates to more traditional concerns of the philosophy of language and mind, and epistemic logic, namely how belief simpliciter does or ought to behave. The paradigmatic theory, probabilism (which holds that degrees of belief ought to satisfy the axioms of probability theory) is given most attention, but competing theories, such as Dempster-Shafer theory, possibility theory, and AGM belief revision theory are also considered. Each of these approaches is represented by one of its major proponents. The papers are specifically written to target advanced undergraduate students with a background in formal methods and beginning graduate students, but they will also serve as first point of reference for academics new to the area.'This is a fascinating collection that brings together issues in traditional and formal epistemology, and succeeds in posing new challenges for both.'Bas C. van Fraassen, San Francisco State University'This is a splendid collection of essays on degrees of belief. [...] the essays are so clear and general that the book also provides an excellent overview of the field for the non-expert.'Frank Arntzenius, University College, Oxford University'Anyone interested in formal epistemology, or in the implications of formal epistemology for philosophy in general, should read this collection.'Brian Weatherson, Rutgers University

In recent years considerable progress has been made in the area of probabilistic graphical models, in particular Bayesian networks and influence diagrams. Probabilistic graphical models have become mainstream in the area of uncertainty in artificial intelligence;contributions to the area are coming from computer science, mathematics, statistics and engineering. This carefully edited book brings together in one volume some of the most important topics of current research in probabilistic graphical modelling, learning from data and probabilistic inference. This includes topics such as the characterisation of conditional independence, the sensitivity of the underlying probability distribution of a Bayesian network to variation in its parameters, the learning of graphical models with latent variables and extensions to the influence diagram formalism. In addition, attention is given to important application fields of probabilistic graphical models, such as the control of vehicles, bioinformatics and medicine.

Our purpose in writing this book was to provide a compendium of stochastic optimizationtechniques,someguidesto wheneachisappropriateinpractical situations, and a few useful ways of thinking about optimization as a p- cess of search in some very rich con?guration spaces. Each of us has come to optimization, traditionally a subject studied in applied mathematics, from a background in physics, especially the statistical physics of random m- tures or materials. One of us (SK) has used ideas developed in the study of magnetic alloys to explore the optimal placement of computer circuits s- ject to many con?icting constraints, while at IBM Research, in Yorktown Heights, NY. The other (JJS) while completing his studies in physics under Prof. Ingo Morgenstern in Regensburg, Germany, and working at the IBM Scienti?c Center Heidelberg, was exposed to optimization problems as varied as scheduling the pickup of fresh milk and planning automobile assembly line schedules. We had the opportunity to work together after SK moved from IBM to a professorship at The Hebrew University of Jerusalem, Israel, and JJS was, for a year, a postdoc there. JJS has taught a course on stochastic optimization at the University of Mainz, where his students have used p- tions of the present manuscript. We hope to make this material readable by undergraduates, and useful to graduate students and practitioners as well, in computer science, applied mathematics, physics, and economics. Mainz, April 2006 JohannesJosefSchneider Jerusalem, April 2006 ScottKirkpatrick Contents Part I Theory Overview of Stochastic Optimization Algorithms 0 General Remarks.........................................

I'm not usually a fan of edited volumes. Too often they are an incoherent hodgepodge of remnants, renegades, or rejects foisted upon an unsuspecting reading public under a misleading or fraudulent title. The volume Scalable Optimization via Probabilistic Modeling: From Algorithms to Applications is a worthy addition to your library because it succeeds on exactly those dimensions where so many edited volumes fail. For example, take the title, Scalable Optimization via Probabilistic M- eling: From Algorithms to Applications. You need not worry that you're going to pick up this book and?nd stray articles about anything else. This book focuseslikealaserbeamononeofthehottesttopicsinevolutionary compu- tion over the last decade or so: estimation of distribution algorithms (EDAs). EDAs borrow evolutionary computation's population orientation and sel- tionism and throw out the genetics to give us a hybrid of substantial power, elegance, and extensibility. The article sequencing in most edited volumes is hard to understand, but from the get go the editors of this volume have assembled a set of articles sequenced in a logical fashion. The book moves from design to e?ciency enhancement and then concludes with relevant applications. The emphasis on e?ciency enhancement is particularly important, because the data-mining perspectiveimplicitinEDAsopensuptheworldofoptimizationtonewme- ods of data-guided adaptation that can further speed solutions through the construction and utilization of e?ective surrogates, hybrids, and parallel and temporal decompositions.

'Mathematics of Uncertainty'provides the basic ideas and foundations of uncertainty, covering the fields of mathematics in which uncertainty, variability, imprecision and fuzziness of data are of importance. This introductory book describes the basic ideas of the mathematical fields of uncertainty from simple interpolation to wavelets, from error propagation to fuzzy sets and neural networks. The book presents the treatment of problems of interpolation and approximation, as well as observation fuzziness which can essentially influence the preciseness and reliability of statements on functional relationships. The notions of randomness and probability are examined as a model for the variability of observation and measurement results. Besides these basic ideas the book also presents methods of qualitative data analysis such as cluster analysis and classification, and of evaluation of functional relationships such as regression analysis and quantitative fuzzy data analysis.