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.

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.

COMIX (configuration of mixing machines) is a system that assists members of the EKATO Sales Department in designing a mixing machine that fulfills the requirements of a customer. It is used to help the engineer design the requested machine and prepare an offer that's to be submitted to the customer. comix integrates more traditional software techniques with explicit knowledge representation and constraint propagation. During the process of routine design, some design decisions have to be made with uncertainty. By including knowledge from process technology and company experience in the mechanical design, a sufficiently high degree of flexibility is achieved that the system can even assist in difficult design situations. The success of the system can be measured by the increase in the quantity and the quality of the submitted offers.

This paper introduces ICET, a new algorithm for cost-sensitive classification. ICET uses a genetic algorithm to evolve a population of biases for a decision tree induction algorithm. The fitness function of the genetic algorithm is the average cost of classification when using the decision tree, including both the costs of tests (features, measurements) and the costs of classification errors. ICET is compared here with three other algorithms for cost-sensitive classification - EG2, CS-ID3, and IDX - and also with C4.5, which classifies without regard to cost. The five algorithms are evaluated empirically on five real-world medical datasets. Three sets of experiments are performed. The first set examines the baseline performance of the five algorithms on the five datasets and establishes that ICET performs significantly better than its competitors. The second set tests the robustness of ICET under a variety of conditions and shows that ICET maintains its advantage. The third set looks at ICET's search in bias space and discovers a way to improve the search.

Temporal difference (TD) methods constitute a class of methods for learning predictions in multi-step prediction problems, parameterized by a recency factor lambda. Currently the most important application of these methods is to temporal credit assignment in reinforcement learning. Well known reinforcement learning algorithms, such as AHC or Q-learning, may be viewed as instances of TD learning. This paper examines the issues of the efficient and general implementation of TD(lambda) for arbitrary lambda, for use with reinforcement learning algorithms optimizing the discounted sum of rewards. The traditional approach, based on eligibility traces, is argued to suffer from both inefficiency and lack of generality. The TTD (Truncated Temporal Differences) procedure is proposed as an alternative, that indeed only approximates TD(lambda), but requires very little computation per action and can be used with arbitrary function representation methods. The idea from which it is derived is fairly simple and not new, but probably unexplored so far. Encouraging experimental results are presented, suggesting that using lambda > 0 with the TTD procedure allows one to obtain a significant learning speedup at essentially the same cost as usual TD(0) learning.

Parti-game is a new algorithm for learning from delayed rewards in high dimensional real-valued state-spaces. In high dimensions it is essential that learning does not explore or plan over state space uniformly. Part i-game maintains a decision-tree partitioning of state-space and applies game-theory and computational geometry techniques to efficiently and reactively concentrate high resolution only on critical areas. Many simulated problems have been tested, ranging from 2-dimensional to 9-dimensional state-spaces, including mazes, path planning, nonlinear dynamics, and uncurling snake robots in restricted spaces. In all cases, a good solution is found in less than twenty trials and a few minutes. 1 REINFORCEMENT LEARNING Reinforcement learning [Samuel, 1959, Sutton, 1984, Watkins, 1989, Barto et al., 1991] is a promising method for control systems to program and improve themselves.

Abstract: There exist a number of negative results ([J), [BR), [KV]) about learning on neural nets in Valiant's model [V) for probably approximately correct learning ("PAClearning"). These negative results are based on an asymptotic analysis where one lets the number of nodes in the neural net go to infinit.y. Hence this analysis is less adequate for the investigation of learning on a small fixed neural net.

Increasing attention has recently been paid to algorithms based on dynamic programming (DP) due to the suitability of DP for learning problems involving control. In stochastic environments where the system being controlled is only incompletely known, however, a unifying theoretical account of these methods has been missing. In this paper we relate DPbased learning algorithms to the powerful techniques of stochastic approximation via a new convergence theorem, enabling us to establish a class of convergent algorithms to which both TD("\) and Q-Iearning belong. 1 INTRODUCTION Learning to predict the future and to find an optimal way of controlling it are the basic goals of learning systems that interact with their environment. A variety of algorithms are currently being studied for the purposes of prediction and control in incompletely specified, stochastic environments. Here we consider learning algorithms defined in Markov environments. There are actions or controls (u) available for the learner that affect both the state transition probabilities, and the probability distribution for the immediate, state dependent costs (Ci(u)) incurred by the learner.

The feature correspondence problem is a classic hurdle in visual object-recognition concerned with determining the correct mapping between the features measured from the image and the features expected by the model. In this paper we show that determining good correspondences requires information about the joint probability density over the image features. We propose "likelihood based correspondence matching" as a general principle for selecting optimal correspondences. The approach is applicable to nonrigid models, allows nonlinear perspective transformations, and can optimally deal with occlusions and missing features.