Intelligence\/sigart Bulletin 

About: Intelligence\/sigart Bulletin is an academic journal. The journal publishes majorly in the area(s): Natural language & Knowledge-based systems. Over the lifetime, 718 publications have been published receiving 14506 citations.

Dyna, an integrated architecture for learning, planning, and reacting

Abstract: Dyna is an AI architecture that integrates learning, planning, and reactive execution. Learning methods are used in Dyna both for compiling planning results and for updating a model of the effects of the agent's actions on the world. Planning is incremental and can use the probabilistic and ofttimes incorrect world models generated by learning processes. Execution is fully reactive in the sense that no planning intervenes between perception and action. Dyna relies on machine learning methods for learning from examples---these are among the basic building blocks making up the architecture---yet is not tied to any particular method. This paper briefly introduces Dyna and discusses its strengths and weaknesses with respect to other architectures.

Cognitive systems based on adaptive algorithms

Abstract: The type of cognitive system (CS) studied here has four basic parts: (1) a set of interacting elementary productions, called classifiers, (2) a performance algorithm that directs the action of the system in the environment, (3) a simple learning algorithm that keeps a record of each classifier's success in bringing about rewards, and (4) a more complex learning algorithm, called the genetic algorithm, that modifies the set of classifiers so that variants of good classifiers persist and new, potentially better ones are created in a provably efficient manner.

Book review: Machine Learning, Neural and Statistical Classification Edited by D. Michie, D.J. Spiegelhalter and C.C. Taylor (Ellis Horwood Limited, 1994)

Abstract: The book is based on the "StatLog" project which lasted from 1990 to 1993, and involved six academic and 6 industrial laboratories. Statisticians, AI researchers in machine learning, and neural net specialists conducted extensive, collaborative work on the task of classification. This well-edited book provides an up-to-date review of different approaches to classification, compares their performance on a wide range of challenging datasets, and draws conclusions on their applicability to realistic industrial problems.

Review of "Problem-Solving Methods in Artificial Intelligence by Nils J. Nilsson", McGraw-Hill Pub.

Abstract: This book is not a survey on theorem proving programs, but the description of a program developed from 1960 to 1965. In the first part there are some generalities on artificial intelligence, and in the second part, some logical explanations, necessary for the comprehension of the program. The third part describes the program. The program has three important features:-it is general. It can study more than one formal system. It receives as data the inference rules and the axioms of the formal system which it must study.-it is capable of invention. It can work without knowing the theorem to be proven. It tries to find interesting theorems; it has only the definition of the interest of a theorem.-it proves theorems, but also metatheorems which are new productions and meta-metatheorems which are new means to get productions. This feature has a great heuristic value. This textbook explains the theoretical ideas underlying problem-solving by heuristically guided, trial-and-error search processes. These search methods are explained by the use of a uniform vocabulary, and several theoretical results about the properties of heuristic search are presented. Several simple example problems, puzzles, and games are used to illustrate the techniques. The author refers to instances in which these same techniques have been successfully applied to problems much more complex than the example problems in his book. The book also includes three chapters that deal with resolution-based theorem-proving in the predicate calculus and its applications to problem solving. Each chapter contains exercises and a section on bibliographical and historical remarks that list some of the more important references related to the subject of the chapter.