Neural networks are processing devices (algorithms) that are loosely modeled after the neuronal structure of the human cerebral cortex but on a much smaller scale. Most neural networks contain some form of ‘learning rule’ which modifies the weights of its internal connections according to the input patterns that it is presented with. In a sense, neural networks learn by example as do their biological counterparts.
Created by: Chris Mitchell
Modified on: Sun, 27 Aug, 2017 at 1:56 PM
Did you find it helpful?Send feedback
Sorry we couldn't be helpful. Help us improve this article with your feedback.