The development of modern kwledge-based systems, for applications ranging from medicine to finance, necessitates going well beyond traditional rule-based programming. Frontiers of Expert Systems: Reasoning with Limited Kwledge attempts to satisfy such a need, introducing exciting and recent advances at the frontiers of the field of expert systems. Beginning with the central topics of logic, uncertainty and rule-based reasoning, each chapter in the book presents a different perspective on how we may solve problems that arise due to limitations in the kwledge of an expert system's reasoner. Successive chapters address (i) the fundamentals of kwledge-based systems, (ii) formal inference, and reasoning about models of a changing and partially kwn world, (iii) uncertainty and probabilistic methods, (iv) the expression of kwledge in rule-based systems, (v) evolving representations of kwledge as a system interacts with the environment, (vi) applying connectionist learning algorithms to improve on kwledge acquired from experts, (vii) reasoning with cases organized in indexed hierarchies, (viii) the process of acquiring and inductively learning kwledge, (ix) extraction of kwledge nuggets from very large data sets, and (x) interactions between multiple specialized reasoners with specialized kwledge bases. Each chapter takes the reader on a journey from elementary concepts to topics of active research, providing a concise description of several topics within and related to the field of expert systems, with pointers to practical applications and other relevant literature. Frontiers of Expert Systems: Reasoning with Limited Kwledge is suitable as a secondary text for a graduate-level course, and as a reference for researchers and practitioners in industry.
Chilukuri Krishna Mohan
Kluwer Academic Publishers
Date of Publication
Computing: Professional & Programming
The Springer International Series in Engineering and Computer Science