Support Vector Learning Machines (SVM) are finding application in pattern recognition, regression estimation, and operator inversion forill-posed problems. Against this very general backdrop, any methods for improving the generalization performance, or for improving the speed in test phase, of SVMs are of increasing interest. Inthis paper we combine two such techniques on a pattern recognition problem. The method for improving generalization performance (the"virtual support vector" method) does so by incorporating known invariances of the problem. This method achieves a drop in the error rate on 10,000 NIST test digit images of 1.4% to 1.0%.

The classes in classification tasks often have a natural ordering, and the training and testing examples are often incomplete. We propose a nonlinear ordinalmodel for classification into ordered classes. Predictive, simulation-based approaches are used to learn from past and classify future incompleteexamples. These techniques are illustrated by making prognoses for patients who have suffered severe head injuries.

We study the effect of noise and regularization in an online gradient-descent learning scenario for a general two-layer student network with an arbitrary number of hidden units. Training examples arerandomly drawn input vectors labeled by a two-layer teacher network with an arbitrary number of hidden units; the examples arecorrupted by Gaussian noise affecting either the output or the model itself. We examine the effect of both types of noise and that of weight-decay regularization on the dynamical evolution ofthe order parameters and the generalization error in various phases of the learning process. 1 Introduction One of the most powerful and commonly used methods for training large layered neural networks is that of online learning, whereby the internal network parameters {J} are modified after the presentation of each training example so as to minimize the corresponding error.

"Emperor AI, Where Is Your New Mind?" The "emperor's new mind" is not here yet, but it is The first two tiers are necessary for developing s AI dead as a science for investigating the Penrose (1989), is the emperor's mind actually an AI theory of the mind simply because without hollow? Should AI be used simply to superficially the right technology, one just could not describe those studies that produce implement such a theory. However, as demonstrated intelligent programs? Several indicators suggest in past AI research and as I argue later, that the answer to these questions might they alone are not sufficient; the machines created be yes.

In adjacent agents from dropping their commitments; (held in Kyoto, Japan, on 10-13 December domains (logical philosophy, social or better, how to regulate 1996). Both the program committee philosophy, decision theory), both legal agents dropping their commitments and the contributors included and social norms have received to a joint action to not disrupt the scientists from different backgrounds considerable, if not satisfactory, attention. The discussion addressed on, has contributed dramatically to These tasks have now entered the several issues: (1) What is the the attention given by the scientific MAS field's common knowledge. Often action is reduced to decision authorization, access regulation, For example, the existence of so-called (that is, a choice among one's privacy maintenance, respect of decency, Georgeff 1991) have shown that we and Tennenholtz 1992). Why? Don't we need a reciprocity.

Intelligence (AAAI), was held 10 to 12 have now become particularly June 1996 in Timberline, Oregon. Finally, we Netherlands, the United States, and would like to acknowledge the support Venezuela. The papers described new of AAAI for student travel funds. Moises Goldszmidt received his Ph.D. in His email address is moises@ Mathematical Institute in Russia.

At a workshop held in Toulouse, France, in 1977, Gallaire, Minker, and Nicolas stated that logic and databases was a field in its own right. This was the first time that this designation was made. The impetus for it started approximately 20 years ago in 1976 when I visited Gallaire and Nicolas in Toulouse, France. In this article, I provide an assessment about what has been achieved in the 20 years since the field started as a distinct discipline. I review developments in the field, assess contributions, consider the status of implementations of deductive databases, and discuss future work needed in deductive databases.

Autonomous agents are computer systems that are capable of independent action in dynamic, unpredictable environments. Agents are also one of the most important and exciting areas of research and development in computer science today. Agents are currently being applied in domains as diverse as computer games and interactive cinema, information retrieval and filtering, user interface design, and industrial process control. Agents '98 will build on the enormous success of the First International Conference on Autonomous Agents (Agents '97), held in Marina del Rey in February 1997. The conference welcomes submissions of original, high quality papers and videos with summaries concerning autonomous agents in a variety of embodiments and playing a variety of roles in their environments.

In this article, we develop a framework for comparing ontologies and place a number of the more prominent ontologies into it. We have selected 10 specific projects for this study, including general ontologies, domain-specific ones, and one knowledge representation system. The comparison framework includes general characteristics, such as the purpose of an ontology, its coverage (general or domain specific), its size, and the formalism used. It also includes the design process used in creating an ontology and the methods used to evaluate it. Characteristics that describe the content of an ontology include taxonomic organization, types of concept covered, top-level divisions, internal structure of concepts, representation of part-whole relations, and the presence and nature of additional axioms. Finally, we consider what experiments or applications have used the ontologies. Knowledge sharing and reuse will require a common framework to support interoperability of independently created ontologies. Our study shows there is great diversity in the way ontologies are designed and the way they represent the world. By identifying the similarities and differences among existing ontologies, we clarify the range of alternatives in creating a standard framework for ontology design.

SEQUITUR is an algorithm that infers a hierarchical structure from a sequence of discrete symbols by replacing repeated phrases with a grammatical rule that generates the phrase, and continuing this process recursively. The result is a hierarchical representation of the original sequence, which offers insights into its lexical structure. The algorithm is driven by two constraints that reduce the size of the grammar, and produce structure as a by-product. SEQUITUR breaks new ground by operating incrementally. Moreover, the method's simple structure permits a proof that it operates in space and time that is linear in the size of the input. Our implementation can process 50,000 symbols per second and has been applied to an extensive range of real world sequences.