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.

Knowledge acquisition tools can be associated with knowledge-based application problems and problem-solving methods. This descriptive approach provides a framework for analysing and comparing tools and techniques, and focuses the task of building knowledge-based systems on the knowledge acquisition process. Knowledge acquisition research strategies discussed at recent Knowledge Acquisition Workshops are shown, distinguishing dimensions of knowledge acquisition tools are listed, and short descriptions of current techniques and tools are given.

Configuring N mutually nonattacking queens on an N-by-N chessboard is a classical problem that was first posed over a century ago. Over the past few decades, this problem has become important to computer scientists by serving as the standard example of a globally constrained problem which is solvable using backtracking search methods. A related problem, placing the N queens on a toroidal board, has been discussed in detail by Poyla and Chandra. Their work focused on characterizing the solvable cases and finding solutions which arrange the queens in a regular pattern. This paper describes a new divide-and-conquer algorithm that solves both problems and investigates the relationship between them.

A major goal of research on networks of neuron-like processing units is to discover efficient learning procedures that allow these networks to construct complex internal representations of their environment. The learning procedures must be capable of modifying the connection strengths in such a way that internal units which are not part of the input or output come to represent important features of the task domain. Several interesting gradient-descent procedures have recently been discovered. Each connection computes the derivative, with respect to the connection strength, of a global measure of the error in the performance of the network. The strength is then adjusted in the direction that decreases the error.

This article is concerned with the computational aspects of combining evidence within the theory of belief functions. It shows that by taking advantage of logical or categorical relations among the questions we consider, we can sometimes avoid the computational complexity associated with brute-force application of Dempster's rule. The mathematical setting for this article is the lattice of partitions of a fixed overall frame of discernment. Different questions are represented by different partitions of this frame, and the categorical relations among these questions are represented by relations of qualitative conditional independence or dependence among the partitions. Qualitative conditional independence is a categorical rather than a probabilistic concept, but it is analogous to conditional independence for random variables.

The paper offers a systematic way of regrouping constraints into hierarchical structures capable of supporting search without backtracking. The method involves the formation and preprocessing of an acyclic database that permits a large variety of queries and local perturbations to be processed swiftly, either by sequential backtrack-free procedures, or by distributed constraint propagation processes.

The automatic interpretation of natural language (in this work, English), database questions formulated by a user untrained in the technical aspects of database querying is an established problem in the field of artificial intelligence. State-of-the-art approaches involve the analysis of queries with syntactic and semantic grammars expressed in phrase structure grammar or transition network formalisms. With such method difficulties exist with the detection and resolution of ambiguity, with the misinterpretation possibilities inherent with finite length look-ahead, and with the modification and extension of a mechanism for other sources of semantic knowledge. This work examines the potential of optimization techniques tomore » solve these problems and interpret natural language, database queries. The proposed method involves developing a 0-1 integer programming problem for each query.

Get Citation Alerts New Citation Alert added!

The ability of learning networks to generalize can be greatly enhanced by providing constraints from the task domain. This paper demonstrates how such constraints can be integrated into a backpropagation network through the architecture of the network. This approach has been successfully applied to the recognition of handwritten zip code digits provided by the U.S. Postal Service. A single network learns the entire recognition operation, going from the normalized image of the character to the final classification.

Artificial intelligence has progressed to the point where multiple cognitive capabilities are being integrated into computational architectures, such as SOAR, PRODIGY, THEO, and ICARUS. Learning in PRODIGY occurs at all decision points and integration in PRODIGY is at the knowledge level; the learning and reasoning modules produce mutually interpretable knowledge structures. Issues in architectural design are discussed, providing a context to examine the underlying tenets of the PRODIGY architecture.