Download AI in the 1980s and Beyond: An MIT Survey by William Eric Leifur Grimson, Ramesh S. Patil PDF

By William Eric Leifur Grimson, Ramesh S. Patil

ISBN-10: 0262071061

ISBN-13: 9780262071062

This selection of essays via 12 individuals of the MIT employees, presents an inside of record at the scope and expectancies of present study in a single of the world's significant AI facilities. The chapters on synthetic intelligence, specialist structures, imaginative and prescient, robotics, and traditional language supply either a large evaluation of present components of job and an overview of the sector at a time of serious public curiosity and fast technological development. Contents: synthetic Intelligence (Patrick H. Winston and Karen Prendergast). KnowledgeBased structures (Randall Davis). Expert-System instruments and strategies (Peter Szolovits). scientific prognosis: Evolution of structures construction services (Ramesh S. Patil). synthetic Intelligence and software program Engineering (Charles wealthy and Richard C. Waters). clever usual Language Processing (Robert C. Berwick). automated Speech reputation and knowing (Victor W. Zue). robotic Programming and synthetic Intelligence (Tomas Lozano-Perez). robotic fingers and Tactile Sensing (John M. Hollerbach). clever imaginative and prescient (Michael Brady). Making Robots See (W. Eric L. Grimson). independent cellular Robots (Rodney A. Brooks). W. Eric L. Grimson, writer of From photographs to Surfaces: A Computational research of the Human Early imaginative and prescient approach (MIT Press 1981), and Ramesh S. Patil are either Assistant Professors within the division of electric Engineering and computing device technological know-how at MIT. AI within the Eighties and past is incorporated within the synthetic Intelligence sequence, edited by way of Patrick H. Winston and Michael Brady.

Show description

Read or Download AI in the 1980s and Beyond: An MIT Survey PDF

Similar intelligence & semantics books

Rough Sets and Intelligent Systems - Professor Zdzislaw Pawlak in Memoriam: Volume 1

This booklet is devoted to the reminiscence of Professor Zdzis{\l}aw Pawlak who passed on to the great beyond nearly six 12 months in the past. he's the founding father of the Polish institution of man-made Intelligence and one of many pioneers in desktop Engineering and laptop technology with around the world effect. He was once a very nice scientist, researcher, instructor and a individual.

Neural Networks and Learning Machines

For graduate-level neural community classes provided within the departments of machine Engineering, electric Engineering, and machine technology.   Neural Networks and studying Machines, 3rd variation is popular for its thoroughness and clarity. This well-organized and fully up to date textual content continues to be the main accomplished remedy of neural networks from an engineering viewpoint.

Principles of Artificial Intelligence

A vintage advent to man made intelligence meant to bridge the space among idea and perform, ideas of man-made Intelligence describes basic AI principles that underlie functions resembling typical language processing, computerized programming, robotics, desktop imaginative and prescient, automated theorem proving, and clever information retrieval.

Extra resources for AI in the 1980s and Beyond: An MIT Survey

Example text

Consider each of those assumptions in turn. Suppose only (a) is false: in that case both X and Yare 6 (by assumption), but ADD-l is not func­ tioning as an adder. We don't know what it is doing, all we know are the symptoms (both inputs are 6, output is 10) . Hence one possibility that is globally consistent with all the symptoms is that ADD-l is broken in the fashion noted. Now suppose only (b) is false. In that case, by assumption, ADD-l is functioning properly and its Y input is 6, and we measured the value at F to be 10.

The fundamental dilemma is the tradeoff between complexity and com­ pleteness. , make no assumptions. But if they do that, they drown in complexity, since they are reduced to exhaustive testing of be­ havior, which is computationally intractable. To handle the complexity we must make some assumptions, yet that eliminates completeness, since it might cause us to overlook the thing that happens to be wrong. The problem is broader than electronic troubleshooting alone. Almost any form of real-world problem solving faces this basic issue.

That's interesting, because now we have another discrepancy, this time between the 6 predicted by simulation as the result of MULT-l, and the 4 inferred from the output observed at F. " ), and Knowledge-Based Systems 33 discover that the problem may be a malfunction in MULT-l ( inputs 3 and 2, output 4) . Finally, suppose only ( c ) is false. In this case ADD-l is working and the value at X is 6 ( by assumption) , the value at F has been measured to be 10, so we infer that the value at Ymust be 4.

Download PDF sample

AI in the 1980s and Beyond: An MIT Survey by William Eric Leifur Grimson, Ramesh S. Patil

by Donald

Rated 4.48 of 5 – based on 40 votes