Qualitative reasoning about physical systems
Qualitative reasoning about physical systems
The :20Brain-state-in-a-box" Neural model is a gradient descent algorithm
Journal of Mathematical Psychology
Learning internal representations by error propagation
Parallel distributed processing: explorations in the microstructure of cognition, vol. 1
| Hi-index | 0.00 |
Neural networks "compute" though not in the way that traditional computers do. It is necessary to accept their weaknesses to use their strengths. We discuss some of the assumptions and constraints that govern operation of neural nets, describe one particular non-linear network---the BSB model---in a little detail, and present two applications of neural network computations to illustrate some of the peculiarities inherent in this architecture. We show how a network can be trained to estimate answers to simple multiplication problems and how a network can be used to disambiguate lexical items by context. In both examples, the way information is represented in the pattern of system unit activities is at least as important as the details of the learning and retrieval algorithms used.