NEURAL NETWORKS
         Neural networks, Expert System and Fuzzy Logic are known as an application areas of A.I. Because of this, I want to give some information about them. First, I am going to begin with Neural Networks.

        The subject of Neural Networks has become increasingly popular, both within the university research environment and in the marketplace in recent years.  It has become quite common to hear references to neural networks in relation to wide range subjects including computer science, neurophyssiology, robotics, automatic control, communications, physics, mathematics, psychology, business ,and many other fields where new applications are being found and investigated.

        Neural networks, or "connectionist techniques", are a novel data processing paradigm with exciting prospects for handling many of the traditionally most challenging computational problems. While advanced techniques such as artificial intelligence have aimed to expand the capabilities of existing by automating human reasoning processes at the software level, neural networks are an attempt to mimics the processing capabilities of the brain at the hardware level - at the level of individual neurons and their overall behavior when combined into large, highly interconnected networks. In doing so, it is hoped that processors can be achieved with many of the most desirable abilities and behaviors of biological systems: highly parallel computation for data-intensive tasks such as vision and speech processing, adaptive learning which is the ability of neural networks to learn how to perform certain tasks by being presented with illustrative examples and self organization of data, generalization, robustness towards noisy or imprecise data, and fault tolerance through distributed or redundant information coding. Such capabilities and performance have been difficult to achieve with traditional computers based on the Von Neumann serial processing paradigm.

    Neural network are being applied in numerous areas involving signal processing and complex pattern recognition and classification tasks, such as image processing, voice and character recognition, noise filtering, data compression, time series prediction and optimization. Another important research area is automatic control, where neural networks are being investigated for the adaptive control of robots and autonomous vehicles, and identification and control of complex or unknown processes. The capacity of neural networks for adaptive learning, non-linear signal mapping and fault tolerance are being exploited to fill many of the persistent gaps present in classical control techniques, and several large scale research and development projects are currently underway to apply neural networks in complex space, industrial, and commercial control applications.