A neural network acquires knowledge through learning.
A neural network's knowledge is stored within inter-neuron connection strengths known as synaptic weights.
The true power and advantage of neural networks lies in their ability to represent both linear and non-linear relationships and in their ability to learn these relationships directly from the data being modeled. Traditional linear models are simply inadequate when it comes to modeling data that contains non-linear characteristics. Models can be made for anything that has historial data that can be backed up.
Subscribe to:
Post Comments (Atom)
True enough. I've been working with NN for years now. They are a rather complex, almost random, system to use. Very useful though in finding process trends. Def post more on this type stuff. Enjoyed the quick definition.
ReplyDelete---------
Justin
TheModernEngineer.blogspot.com