Computer Science/Neural Networks
From Wikibooks, the open-content textbooks collection
[edit] Introduction
Neural Networks are a simulation of biological neural networks (one example of a biological neural network is a brain) and are a radical approach to computational problem solving. Neural networks in computing are often referred to as Artificial Neural Networks, or ANNs. There are different types of these networks, and this book will take the reader through the general subject.
[edit] Beginning
Principle The principle behind a neural network is that you have some set of nodes N, where each element(n) of N receives a certain input (i) from all other nodes that connect to it, it then sums the weight of its inputs and that is its value, thus (v(n) = i1 + i2 + i3 + ... + ip); further if T is met then n(x) will send its V to another node n(y) where V = (v(n(x)) x w(n(X),n(y))) where w is the weight (i.e. strength of connection) between n(x) and n(y). Among the set N there is a subset denoted I such that the value of I comes from an external source and is thus called the input nodes, and there will be a subset O where its members will deliver its values to an external source and are thus called output nodes.
