Artificial neural networks are developed by programming. The programming mimics the properties of biological neurons in the nervous system. It uses artificial intelligence and cognitive modeling that try to imitate the biological neural networks. The advantages of artificial neural networks are it can perform several tasks that a linear program can't, if one part of the artificial neural network does not work, it can continue without any problems, it can learn and does not need to be reprogrammed, it can be implemented in any kind of application, and it can be implemented without any problems. The disadvantages of artificial are that neural network needs a lot of training to use, the architecture of a artificial neural network is a lot different than the architecture of common microprocessors so it needs to be emulated, and also it requires high processing time for large neural networks.
They are the newest signal processing technologies and they have two main functions which are pattern classifiers and non linear adaptive filters. It is an adaptive technology. Every parameter is changed when it is run and it is used to solve the problem in a matter. It uses step-by-step procedures that optimize a criteria that it places. It is very flexible because it is non linear.
Artificial intelligence has been applied to image analysis, speech recognition, and adaptive control. They use this especially for video games and autonomous robots. They use this for car building companies whose robots builds cars more productively for them. Artificial intelligence allows the robots to act in a way where it normally would not be able to act.
The second part is the cognitive modeling. In cognitive modeling, it involves the physical or mathematical modeling of the behavior of neural systems.
One example of a neural network application is bankruptcy prediction
One example of a neural network application is bankruptcy prediction