K. Assiter, S. Best, B. Faidell, and C. Weaver (USA)
Modeling, Computational Neuroscience, Simulation
Computational neuroscientists develop models for representing the structure and behavior of in vivo neural networks. These models include the short term and long term learning equations for classical conditioning [1, 2], the additive network equations, the feed-forward competitive shunting networks, the Hopfield network [3] and the recurrent competitive networks. Tools have been developed from these (and other) established models, but, in general, it is hard to find tools that a researcher can use to explore new paradigms of in vivo neural network learning. We propose a visualization system which will facilitate the development of new neural network paradigms. NeuralModeler will provide a flexible environment for building and simulating neural models: it will be adaptable to any realizable neural model; it will allow the re-use of constructed neural components; it will be portable to run on any machine; it will be easy to use and learn, and it will allow the use of dynamic real-time stimuli input.
Important Links:
Go Back
