What is an Artificial Neural Network?

 

The human brain is a neural network; it consists of millions or nerve cells (neurons) which take in inputs and send outputs, via synapses, to other nerve cells. This allows the brain to make predictions, or carry out functions, such as breathing, or holding your breath when under water. The most important element of a neural network is that it learns. It starts out with a simple premise and some inputs, and continually tests the outputs against reality, amending the way it processes data, until it is able to make predictions with a high degree of accuracy.

 

An artificial neural network is a computer representation of the humaun brain. It has inputs and outputs, and in the middle, one or more "hidden" layers, that act as neurons making decisions.

 

Both natural and artificial neural networks can learn in a number of ways: supervised, unsupervised or reinforcement learning; whichever method is used the network ends up being better at making decisions than is started.

 

Click here for our next Disruptive Technology Explained blog on Deep Learning.