We study the process of multi-agent reinforcement learning in the context ofload balancing in a distributed system, without use of either centralcoordination or explicit communication. We first define a precise frameworkin which to study adaptive load balancing, important features of which are itsstochastic nature and the purely local information available to individualagents. Given this framework, we show illuminating results on the interplaybetween basic adaptive behavior parameters and their effect on systemefficiency. We then investigate the properties of adaptive load balancing inheterogeneous populations, and address the issue of exploration vs.exploitation in that context. Finally, we show that naive use ofcommunication may not improve, and might even harm system efficiency.

We present algorithms that learn certain classes of function-free recursive logic programs in polynomial time from equivalence queries. In particular, we show that a single k-ary recursive constant-depth determinate clause is learnable. Two-clause programs consisting of one learnable recursive clause and one constant-depth determinate non-recursive clause are also learnable, if an additional ``basecase'' oracle is assumed. These results immediately imply the pac-learnability of these classes. Although these classes of learnable recursive programs are very constrained, it is shown in a companion paper that they are maximally general, in that generalizing either class in any natural way leads to a computationally difficult learning problem. Thus, taken together with its companion paper, this paper establishes a boundary of efficient learnability for recursive logic programs.

In a companion paper it was shown that the class of constant-depth determinate k-ary recursive clauses is efficiently learnable. In this paper we present negative results showing that any natural generalization of this class is hard to learn in Valiant's model of pac-learnability. In particular, we show that the following program classes are cryptographically hard to learn: programs with an unbounded number of constant-depth linear recursive clauses; programs with one constant-depth determinate clause containing an unbounded number of recursive calls; and programs with one linear recursive clause of constant locality. These results immediately imply the non-learnability of any more general class of programs. We also show that learning a constant-depth determinate program with either two linear recursive clauses or one linear recursive clause and one non-recursive clause is as hard as learning boolean DNF. Together with positive results from the companion paper, these negative results establish a boundary of efficient learnability for recursive function-free clauses.

The Association for the Advancement of Artificial Intelligence held its 1994 Fall Symposium Series on November 4-6 at the Monteleone Hotel in New Orleans, Louisiana. This article contains summaries of the five symposia that were conducted: (1) Control of the Physical World by Intelligent Agents, (2) Improving Instruction of Introductory AI, (3) Knowledge Representation for Natural Language Processing in Implemented Systems, (4) Planning and Learning: On to Real Applications, and (5) Relevance.

Countrywide loan-underwriting expert system (clues) is an advanced, automated mortgage-underwriting rule-based expert system. The system was developed to increase the production capacity and productivity of Countrywide branches, improve the consistency of underwriting, and reduce the cost of originating a loan. The system receives selected information from the loan application, credit report, and appraisal. It then decides whether the loan should be approved or whether it requires further review by a human underwriter. If the system approves the loan, no further review is required, and the application is funded. clues has been in operation since February 1993 and is currently processing more than 8500 loans each month in over 300 decentralized branches around the country.

's central purpose is to find the most The use of composite materials, Composite part fabrication requires two especially in aerospace applications, is on the major steps: layup and curing. Layup is the increase because of their unique weight and painstaking process in which multiple layers strength qualities. Depending on the orientation of graphite and fiberglass composite material of the graphite fibers, a part can be are fitted by hand on the exterior of a contoured extremely flexible in one direction but rigid mold. In addition, a part made from takes from two to seven days, depending on composite material is both lighter and the size of the mold and the skill of the technician. The increased use of In the second step, curing, the molded graphite parts, as well as the high cost of a composite material is hardened by pressurized spoiled part (as much as $50,000 for a single heating in a large convection autoclave.

The vertical launch system (vls) tech-assist expert system is being used by the in-service engineering agent as a force multiplier to maintain the readiness, with fewer resources, of a growing population of vlss in the U.S. Navy fleet. This article describes the collaborative development of this knowledge-based system for diagnosis; its main features, including case-based and model-based reasoning; and the lessons we learned from the process.

Interactive simulation environments constitute one of today's promising emerging technologies, with applications in areas such as education, manufacturing, entertainment, and training. These environments are also rich domains for building and investigating intelligent automated agents, with requirements for the integration of a variety of agent capabilities but without the costs and demands of low-level perceptual processing or robotic control. Our project is aimed at developing humanlike, intelligent agents that can interact with each other, as well as with humans, in such virtual environments. Our current target is intelligent automated pilots for battlefield-simulation environments. These dynamic, interactive, multiagent environments pose interesting challenges for research on specialized agent capabilities as well as on the integration of these capabilities in the development of "complete" pilot agents. We are addressing these challenges through development of a pilot agent, called TacAir-Soar, within the Soar architecture. This article provides an overview of this domain and project by analyzing the challenges that automated pilots face in battlefield simulations, describing how TacAir-Soar is successfully able to address many of them -- TacAir-Soar pilots have already successfully participated in constrained air-combat simulations against expert human pilots -- and discussing the issues involved in resolving the remaining research challenges.

Hughes Missile Systems (Tucson, Lear Astronics (Santa Monica and to "enter" the surgical area, as if they Ariz.) is providing intelligent character Ontario, Calif.) is combining neural were actually there. Cross/Blue Shield (New York, N.Y.) to enhance its Autonomous Landing FuziWare (Knoxville, Tenn.), a developer expedite the processing of medical Guidance (ALG) system. Empire will install the is using a neural network-based tools for business and engineering ICRs at its Yorktown Heights and massively parallel coprocessor for solutions, has received a patent from Manhattan offices, where they will be real-time image processing in the the U.S. Department of Commerce ALG system, which enables commercial Patent and Trademark Office for its used to process about 10,000 documents and military aircraft pilots to FuziCalc product, a fuzzy spreadsheet per day. The claims in the patent cover various fuzzy number The Boston Museum of Fine Arts Researchers at Georgia Tech (Atlanta, interface elements as well as the (Boston, Mass.) has developed a virtual Ga.) have created intelligent agent entire fuzzy number processing system. TOAK navigates and surgical equipment, has implemented through multiple networks and a virtual reality application a complex 3D model derived from across diverse computer systems to for technical design presentation.

The Workshop on Using Knowledge in Its Context was held in Chambery, France, on 28 August 1993, preceding the Thirteenth International Joint Conference on Artificial Intelligence (IJCAI-93). This article provides a summary of the discussions between the participants before (by e-mail) and during the one-day workshop. It is clear from these discussions that the notion of context is far from defined and is dependent in its interpretation on a cognitive science versus an engineering (or system building) point of view. In identifying the two points of view, this workshop permitted us to go one step further than previous workshops (notably Maskery and Meads [1992] and Maskery, Hopkins, and Dudley [1992]). Once a distinction is made on the viewpoint, one can achieve a surprising consensus on the aspects of context that the workshop addressed -- mainly, the position, the elements, the representation, and the use of context. Despite this consensus on the aspects of context, agreement on the definition of context was not yet achieved.