Hutchins not only presents machine translation research (such as problems of machine translation It is the theories, algorithms, and designs practical versus theoretical, empirical also not clear that the AI philosophy but also the history, goals, assumptions, versus perfectionist, and direct versus of understanding and meaning (p 327) and constraints of each project.

This article explores the premise that a formalized representation of empirical studies can play a central role in computer- based decision support. The specific motivations underlying this research include the following propositions: (1) Reasoning from experimental evidence contained in the clinical literature is central to the decisions physicians make in patient care. (2) A computational model based on a declarative representation for published reports of clinical studies can drive a computer program that selectively tailors knowledge of the clinical literature as it is applied to a particular case. (3) The development of such a computational model is an important first step toward filling a void in computer-based decision support systems. Furthermore, the model can help us better understand the general principles of reasoning from experimental evidence both in medicine and other domains. Roundsman is a developmental computer system that draws on structured representations of the clinical literature to critique plans for the management of primary breast cancer. Roundsman is able to produce patient-specific analyses of breast cancer-management options based on the 24 clinical studies currently encoded in its knowledge base. The Roundsman system is a first step in exploring how the computer can help bring a critical analysis of the relevant literature, structured around a particular patient and treatment decision, to the physician.

The notion of generic design, although it has been around for 25 years, is not often articulated; such is especially true within Newell and Simon's (1972) information-processing theory (IPT) framework. Design is merely lumped in with other forms of problem-solving activity. Intuitively, one feels there should be a level of description of the phenomenon that refines this broad classification by further distinguishing between design and nondesign problem solving. However, IPT does not facilitate such problem classification. This article makes a preliminary attempt to differentiate design problem solving from nondesign problem solving by identifying major invariants in the design problem space.

Hutchins not only presents machine translation research (such as problems of machine translation It is the theories, algorithms, and designs practical versus theoretical, empirical also not clear that the AI philosophy but also the history, goals, assumptions, versus perfectionist, and direct versus of understanding and meaning (p 327) and constraints of each project.

A panel session at the 1989 International Joint Conference on artificial intelligence in Los Angeles dealt with the subject of knowledge-based systems; the session was entitled "Expert Systems: How Far Can They Go?" The panelists included Randall Davis (Massachusetts Institute of Technology); Stuart Dreyfus (University of California at Berkeley); Brian Smith (Xerox Palo Alto Research Center); and Terry Winograd (Stanford University), chairman. The article begins with Winograd's original charge to the panel, followed by lightly edited transcripts of the panel's remarks. Part 1 includes presentations from Winograd and Dreyfus. Part 2, which will appear in the Summer 1989 issue, includes presentations from Smith and Davis and concludes with the panel discussion. Although three years have passed since this session took place, the issues raised and the points discussed are no less relevant today.

Hutchins not only presents machine translation research (such as problems of machine translation It is the theories, algorithms, and designs practical versus theoretical, empirical also not clear that the AI philosophy but also the history, goals, assumptions, versus perfectionist, and direct versus of understanding and meaning (p 327) and constraints of each project.

Decision theory and nonmonotonic logics are formalisms that can be employed to represent and solve problems of planning under uncertainty. We analyze the usefulness of these two approaches by establishing a simple correspondence between the two formalisms. The analysis indicates that planning using nonmonotonic logic comprises two decision-theoretic concepts: probabilities (degrees of belief in planning hypotheses) and utilities (degrees of preference for planning outcomes). We present and discuss examples of the following lessons from this decision-theoretic view of nonmonotonic reasoning: (1) decision theory and nonmonotonic logics are intended to solve different components of the planning problem; (2) when considered in the context of planning under uncertainty, nonmonotonic logics do not retain the domain-independent characteristics of classical (monotonic) logic; and (3) because certain nonmonotonic programming paradigms (for example, frame-based inheritance, nonmonotonic logics) are inherently problem specific, they might be inappropriate for use in solving certain types of planning problems. We discuss how these conclusions affect several current AI research issues.

A class of connectionist networks is described that has learned to play backgammon at an intermediate-to-advanced level. The networks were trained by back-propagation learning on a large set of sample positions evaluated by a human expert. In actual match play against humans and conventional computer programs, the networks have demonstrated substantial ability to generalize on the basis of expert knowledge of the game. This is possibly the most complex domain yet studied with connectionist learning. New techniques were needed to overcome problems due to the scale and complexity of the task.

ABSTRACT Classifier systems are massively parallel, message-passing, rule-based systems that learn through credit assignment (the bucket brigade algorithm) and rule discovery (the genetic algorithm). They typically operate in environments that exhibit one or more of the following characteristics: (1) perpetually novel events accompanied by large amounts of noisy or irrelevant data; (2) continual, often real-time, requirements for action; (3) implicitly or inexactly defined goals; and (4) sparse payoff or reinforcement obtainable only through long action sequences. Classifier systems are designed to absorb new information continuously from such environments, devising sets of compet- ing hypotheses (expressed as rules) without disturbing significantly capabilities already acquired. This paper reviews the definition, theory, and extant applications of classifier systems, comparing them with other machine learning techniques, and closing with a discussion of advantages, problems, and possible extensions of classifier systems. Artificial Intelligence, 40 (1-3), 235-82.

In IJCAI-89, pp. 334–340.