Research into the computational capabilities and limitations of neural networks is fundamental to understanding information processing in the brain.
Biologically based neural network models have proven important for integrating biophysical, physiological, and anatomical information for performing experiments (simulations) that currently are not feasible for experimentalists. Artificial neural networks have proven utility in their own right, providing tools for nonlinear signal processing and pattern classification, as well as providing a mathematical framework for gathering insight into a more abstract level of brainlike computation. The challenge for the future will be to develop a tighter integration between the computational framework and neurobiology, while at the same time incorporating molecular and genomic data for discovering the basic principles governing brain function.
Was this article helpful?