Advocates and critics of AI have long engaged in a debate that has generated a great deal of heat but little light. Whatever the merits of specific contributions to this ongoing debate, the fact that it continues points to the need for a reflective examination of the foundations of AI by its active practitioners. Following the lead of Earl MacCormac, we hope to advance such a reflective examination by considering questions of metaphor in science and the computational metaphor in AI. Specifically, we address three issues: the role of metaphor in science and AI, an examination of the computational metaphor, and an introduction to the possibility and potential value of using alternative metaphors as a foundation for AI theory.

Fields 3-8 of table 1 of the survey and general results, a discussion represent purposes, specifically, to define of the four hypotheses, and two sections models (field 3), prove theorems about the at the end of the article that contain details of models (field 4), present algorithms (field 5), the survey and statistical analyses. The next analyze algorithms (field 6), present systems section (The Survey) briefly describes the 16 or architectures (field 7), and analyze them substantive questions I asked about each (field 8). These purposes are not mutually paper. One of the closing sections (An Explanation exclusive; for example, many papers that of the Fields in Table 1) discusses the present models also prove theorems about criteria for answering the survey questions the models.

Thus, traditional (deterministic) planners can be useful in probabilistic domains.

The VLSI cell placement problem is known to be NP-complete. This paper presents a survey of the various approaches and techniques for this problem. It also gives a comprehensive tutorial on the subject, providing an excellent introduction to the terminology and classification of placement algorithms. With the growing diversity of the terms appearing in the literature, I found the explicit warning about synonymous usage of words like module, cell, and element or net, wire, and interconnect to be helpful. The placement algorithms whose emphasis is on standard cell and macro placement fall into five groups, according to their underlying technique: (1) simulated annealing, (2) force-directed, (3) minimum-cut, (4) numerical optimization, and (5) evolution based. The origins of the first two are in physical laws.

It is well known that for many NPcomplete problems, such as K-Sat, etc., typical cases are easy to solve; so that computationally hard cases must be rare (assuming P NP). This paper shows that NPcomplete problems can be summarized by at least one "order parameter", and that the hard problems occur at a critical value of such a parameter.

One of the most crucial problems in any computer system that involves representing the world is the representation of time. This includes applications such as databases, simulation, expert systems, and applications of Artificial Intelligence in general. In this brief article, I will give a survey of the basic techniques available for representing time, and then talk about temporal reasoning in a general setting as needed in AI applications. Quite different representations of time are usable depending on the assumptions that can be made about the temporal information to be represented. Can one assume that a timestamp can be assigned to each event, or barring that, that the events are fully ordered?

Based on the experience in manufacturing production scheduling problems which the AI community has amassed over the last ten years, a workshop was held to provide a forum for discussion of the issues encountered in the design of AI-based scheduling systems. Several topics were addressed including: the relative virtues of expert system, deep method, and interactive approaches, the balance between predictive and reactive components in a scheduling system, the maintenance of convenient scheduling descriptions, the application of the ideas of chaos theory to scheduling, the state of the art in schedulers which learn, and the practicality and desirability of a set of benchmark scheduling problems. This article expands on these issues, abstracts the papers which were presented, and summarizes the lengthy discussions that took place.

This symposium was very successful and was perhaps the most unusual of the spring symposia this year. It brought together for the first time distinguished researchers from many diverse disciplines to discuss and share results on a particular topic of mutual interest. The disciplines included machine learning, computational learning theory, computer vision, pattern recognition, perceptual psychology, statistics, information theory, theoretical computer science, and molecular biology, with the involvement of the latter group having lead to a joint session with the AI and Molecular Biology symposium.

It is generally admitted that generalization performance of back-propagation networks (Rumelhart, Hinton & Williams, 1986) will depend on the relative size ofthe training data and of the trained network. By analogy to curve-fitting and for theoretical considerations, the generalization performance of the network should decrease as the size of the network and the associated number of degrees of freedom increase (Rumelhart, 1987; Denker et al., 1987; Hanson & Pratt, 1989). This paper examines the dynamics of the standard back-propagation algorithm (BP) and of a constrained back-propagation variation (CBP), designed to adapt the size of the network to the training data base. The performance, learning dynamics and the representations resulting from the two algorithms are compared.