Learning internal representations by error propagation
Parallel distributed processing: explorations in the microstructure of cognition, vol. 1
Selection of a neural network system for visual inspection
ACM SIGSIM Simulation Digest
Hi-index | 0.00 |
Parallel 'neural' algorithms such as back-propagation[8] have attracted a great deal of interest because of their apparent potential for learning complex pattern recognition tasks. However the basic algorithms have no temporal components, although many of the problems to which they could be applied do have temporal aspects. One strategy for introducing temporal processing has been proposed by Jordan[7] for modeling articulation. A related architecture has been employed by Allen[1,3,4] for modeling a variety of high-level cognitive processes.