ARTIFICIAL NEURAL NETWORKS

SEMINAR ON ARTIFICIAL NEURAL NETWORKS : Most people when asked if they think that computers can never become sentient respond quickly to anything, and suggest that computers can not learn. However, neural networks seem to do. Neural networks are a variety of computer models parts of the base set of simple functions. Inspired style of computation in biological systems, a neural network can be considered Because the composition of simple, interconnected slow (neurons) are in parallel, communicating with each other by way connections. Neural networks are different from the other computer and mathematical techniques for design reasons. They are the management of devices that can current algorithms, or hardware, which is modeled after operation human brain. Most neural networks have some sort of "training" rule The connection weights are adjusted on the basis of present models. In other words, the neural networks 'learn' examples, such as children learn to recognize dogs from examples of dogs and have some structural capacity of generalization. The most important aspects of neural networks is that they allow computer to learn, and they have the potential for parallelism. This means that allow the computer to solve several problems at once. Neural networks can perform a variety of tasks as any regular computer. They are particularly useful in data processing problems where the input do not follow strict rules strictly, but rather a global trend. Neural Network has applications in fields as diverse as the interpretation, forecasting, diagnosis, planning, monitoring, development, restoration, education, supervision, categorization and pattern recognition. Therefore, the neural network is a exponential growth in the field of real-time applications in the new era.

No comments:

Post a Comment