A comparison of training techniques: ADALINE, back progagation and genetic algorithms

This study is in the area of neural network architectures and training algorithms. The emphasis is placed on the comparisons of ADALINE, Back Propagation and Genetic Algorithms in training neural networks. Concrete examples are developed to illustrate and reveal the fundamental theories of neural networks, and demonstrate the strengths and weaknesses of ADALINE, Back Propagation and Genetic Algorithms. An object-oriented approach is applied in the overall analysis and design of the neural network architectures. The Object-oriented programming with C++ is used to facilitate the development and implementation of the neural network architectures and training algorithms.