A neural network is a series of algorithms that endeavors to recognize basic relationships in a set of record through a process that mimics the way the human brain operates. In this method, neural networks defines systems of neurons, either organic or artificial.
Neural Networks are analytic techniques modeled after the (hypothesized) processes of learning in the cognitive system and the neurological functions of the brain and capable of predicting new observations (on specific variables) from other observations after implementing a process of so-called learning from existing information. Neural Networks are one of the Data Mining techniques.
The first phase is to design a specific network architecture (that involves a definite number of “layers” each including a specific number of “neurons”). The size and structure of the network need to match the nature (e.g., the formal complexity) of the investigated phenomenon. Because the latter is not known very well at this early stage, this task is not easy and often involves multiple “trials and errors.”
The new network is then subjected to the process of “training.” In that phase, neurons apply an iterative process to the number of inputs (variables) to adjust the weights of the network to optimally predict (in traditional terms one could say, find a “fit” to) the sample data on which the “training” is performed. After the phase of learning from an existing data set, the new network is ready and it can then be used to generate predictions.
Neural networks have view an eruption of interest over the last few years, and are being successfully used across an extraordinary area of problem domains, in areas as diverse as finance, medicine, engineering, geology, and physics. There are two elements of a neural network which are as follows −
Power − Neural networks are very refined modeling techniques adequate of modeling extremely complex functions. In particular, neural networks are nonlinear. For some years linear modeling has been the generally used methods in most modeling domains because linear models have well-known optimization strategies.
Ease of use − Neural networks learn by example. The neural network user gathers representative data and then invokes training algorithms to automatically learn the structure of the data.
Although the user does required to have some heuristic knowledge of how to choose and prepare records, how to choose an appropriate neural network, and how to execute the results, the level of user knowledge needed to successfully use neural networks is much lower than would be the case using (for instance) some more traditional nonlinear statistical methods.