Artificial neural networks

Artificial neural network makes deep learning possible. They are adaptive systems that are inspired by the functioning processes of the human brain. They're systems that are able to modify their internal structure in relation to a function objective.

ANNs are high in pattern recognition-like abilities, which are needed for pattern recognition and decision making are and are robust classifiers with the ability to generalise and make decisions from large and somewhat fuzzy input data.

The most typical problem that an ANN can deal with can be expressed as follows: given N variables, about which it is easy to gather data, and M variables, which differ from the first and about which it is difficult and costly to gather data, assess whether it is possible to predict the values of the M variables on the basis of the N variables.

What are the basic components of an ANN?
They have several computing units called neurons, nodes or processing elements, which are organised in three connected layers: the input layer, hidden layers and output layer. Nodes receive inputs, process them using a function and produce an output. Connections link nodes together, and each connection has a weight that determines how strongly it influences the nodes it connects. Positive weights indicate excitation, while negative weights indicate inhibition. The network learns by adjusting these connection weights over time.

How does an ANN learn?
An ANN learns through a process where it modifies its connections over time based on the data it receives. This is called the 'Law of Learning.' The network adjusts the weights of its connections to better understand the relationships within the data. This learning process allows the ANN to adapt to the data structure and, in effect, 'understand' the environment it's operating in.

When to use ANNs. ANNs are useful when you have a tricky problem, other ways of solving it haven't worked well, and getting close to the right answer is good enough. They're also great if you have lots of examples of the problem and solving it can save time or money.