The artificial neural network is a new interdisciplinary science. It is also one of the major contents of nonlinear science research. The aim of this course is to help students master the basic concept, principles and the basic models about the artificial neural network, and solve practical problems combining with their specialty and engineering applications, and understand the attractive developmental trends of the artificial neural network.
The main contents include: (1) The fundamental concepts of the artificial neural network - neurons and its mechanisms of information processing, connection weights, activation functions, learning rate parameters and so on. (2) the model of the perceptron and its learning rules. (3) the feedforward neural network, in particular, BP model, its design and the influences of the initial connection weights. (4) the principle and method of continuous and discrete Hopfield feedback neural network models. (5) the basic principles and methods of competitive networks and self-organizing neural networks, mainly include self-organizing feature maps, learning vector quantization and the adaptive resonance theory networks. (6) the basic concepts, algorithm of genetic algorithm and its applications in artificial neural networks. (7) the basic concepts and principles of stochastic neural network.