Artificial neural networks (ANN) is mathematical models and their software and hardware implementation, based on the principle of functioning of biological neural networks – networks of nerve cells of a living organism. Systems, architecture, and principles are based on the analogy with the brain of living beings. A key element of these systems is the artificial neuron as a simulation model of nerve cells in the brain, i.e., a biological neuron. The term originated in the study of the processes that occur in the brain, and in attempt to simulate these processes. The first such attempt was made by McCulloch and Pitts with their neural networks. As a result, after the development of learning algorithms, derived models have been used for practical purposes: in forecasting problems, for pattern recognition, in control problems and others.
ANN represents a system of interconnected and interacting simple processing units (artificial neurons). These processors are usually fairly simple, especially compared to the processors used in personal computers. Each processor has a similar online deal only with signals that it receives from time to time, and signals that it periodically sends to the other processors. Nevertheless, being connected in a very large network with controlled interaction, these simple processors capable to cope with most challenges. From the perspective of machine learning, neural network is a special case of pattern recognition methods, discriminant analysis, clustering methods, etc.
From a mathematical point of view, neural networks learning is a multivariable nonlinear optimization problem. In terms of cybernetics, neural network is used in problems of adaptive control and algorithms for robotics. In terms of computing and programming, neural network is intended to solve the problem of effective parallelism. And in terms of artificial intelligence, ANN is the basis of the philosophy of the connectionism and the main direction in the structural approach to study the possibility of building (modeling) of natural intelligence with computer algorithms.
Neural networks are not programmed in the usual sense of the word, they learn. The opportunity to learn is one of the main advantages of neural networks over traditional algorithms. Technically, the learning implies finding the coefficients of connections between neurons. During learning, the neural network is able to detect complex relationships between inputs and output, and perform generalization. This means that in case of successful learning network will return a true result, based on data that were missing from the study sample, as well as parent and / or “noisy,” partly false data.
To write a good research proposal on artificial neural networks, you should consult free sample research paper topics on the subject, which will teach you that understanding the function of the neuron and the pattern of its connections allowed researchers to create mathematical models to test their theories.
Are you looking for a top-notch custom research paper on Artificial Neural Network topics? Is confidentiality as important to you as the high quality of the product?
Try our writing service at EssayLib.com! We can offer you professional assistance at affordable rates. Our experienced PhD and Master’s writers are ready to take into account your smallest demands. We guarantee you 100% authenticity of your paper and assure you of dead on time delivery. Proceed with the order form:
Please, feel free to visit us at EssayLib.com and learn more about our service!