what is a neural network

Our brains are biological neural networks. A major function of the brain is to process data and information that we take in so that we can make decisions in the light of our training and experiences.

Artificial neural networks work on principles similar to some of the processing that occurs in the brain. Neural networks are trained on numerous examples of data relevant to the decision or prediction that they are required to produce. When trained and tested, a neural network is then able to make decisions, judgments and predictions on relevant new data that is presented to it.

There are several different types of neural networks. The type described here is a back propagation feed forward model. It is trained on past data from which it can project results forward. For example data used by expert recruitment staff to determine whether or not a new applicant is suitable for a particular job could be used in conjunction with data reflecting how well previous applicants had done, to train a neural network to make sound decisions on the suitability of each new candidate.

Neural networks in this category are in current use to make quality decisions in many areas where decision making is complex and difficult to achieve.