Book review: Connectionist Symbol Processing edited by Geoffrey Hinton (MIT/Elsevier Press) and High-Level Connectionist Model edited by John A. Barnden and Jordan B. Pollack (Advances in Connectionist and Neural Computation Theory Volume 1 John A. Barnden, series editor Ablex Publishing Corporation)

Authors:
A. David Redish
Affiliations:
School of Computer Science Carnegie Mellon University David.Redish@cs.cmu.edu
Venue:
ACM SIGART Bulletin
Year:
1992

Citing 2
Cited 0

On learning the past tenses of English verbs

Parallel distributed processing: explorations in the microstructure of cognition, vol. 2
A distributed neural network architecture for hexapod robot locomotion

Neural Computation

Quantified Score

Hi-index	0.00

Visualization

Abstract

Some people see the AI community as split into two camps, the connectionist camp, which advocates massively parallel neural networks, and the symbolic camp, which insists that intelligence requires symbols. Both of these books accept this split and try to bridge it. Ever since Minsky and Papert's book, Perceptrons [6] which showed that single layer linear threshold units could not calculate parity, those in the connectionist camp have bristled at the success, albeit limited, of the symbolic camp.