RAM-based neural networks are designed to be efficiently implemented in hardware. The desire to retain this property influences the training algorithms used, and has led to the use of reinforcement (reward-penalty) learning. An analysis of the reinforcement algorithm applied to RAM-based nodes has s
A continuous input RAM-based stochastic neural model
β Scribed by Denise Gorse; John G. Taylor
- Publisher
- Elsevier Science
- Year
- 1991
- Tongue
- English
- Weight
- 788 KB
- Volume
- 4
- Category
- Article
- ISSN
- 0893-6080
No coin nor oath required. For personal study only.
π SIMILAR VOLUMES
A model of neural processing is proposed which is able to incorporate a great deal of neurophysiological detail, including effects associated with the mechanics of postsynaptic summation and cell surface geometry and is capable of hardware realisation as a 'probabilistic random access memory' (pRAM)
This paper deals with stochastic m-compartmental systems with continuous timedependent infusions into all compartments and reversible time-independent flows between any two compartments. A methodology for the first two moments of the distribution of the number of units in the different compartments
The continuous network design problem (CNDP) is known to be difficult to solve due to the intrinsic properties of non-convexity and nonlinearity. Such kinds of CNDP can be formulated as a bi-level programme, in which the upper level represents the designer's decisions and the lower level the travell