A CONTINUOUS INPUT RAM-BASED STOCHASTIC NEURAL MODEL.
657 - 665.
An extension of the probabilistic random access memory (pRAM) neural model is presented, which is shown to have a natural capacity for generalisation. This is displayed in one-and two-dimensional spatial learning tasks, using a form of reinforcement training. The model is then further extended to allow for the learning of temporal sequences, and this capacity is demonstrated in a simple temporal learning problem.
|Title:||A CONTINUOUS INPUT RAM-BASED STOCHASTIC NEURAL MODEL|
|Keywords:||RAMS, PRAMS, STOCHASTIC MODELS, GENERALIZATION, REINFORCEMENT, SPATIOTEMPORAL LEARNING, NETWORKS, NETS|
|UCL classification:||UCL > School of BEAMS > Faculty of Engineering Science
UCL > School of BEAMS > Faculty of Engineering Science > Computer Science
Archive Staff Only