基于ART理论的自组织神经网络模型在水资源分类中的应用

BP和RBF神经网络技术以其强大的学习功能应用于水资源分类 ,取得了很好的效果。但当不具备已知样本时 ,以上技术很难应用。提出了可塑性较强、无监督的A -K网络模型 ,阐述了其基本原理和算法 ,并将其用于水文水资源研究领域中。实例表明 ,该方法能较理想地解决已知样本的分类问题 ,具有良好的应用前景

THE APPLICATION OF SELF-ORGANIZING NEURAL NETWORK MODELS BASED ON THE THEORY OF ART IN WATER RESOURCES CLASSIFICATION

Back propagation and radial basis function neural network methods have been applied to water resources areas due to theirs powerful learning abilitys and many good results have been achieved. These methods can not be applied successfully if there are no known samples for learning. In this paper, the model of A-K neural network, which has plasticity, and is unsupervised, is worked out. The principle and algorithm are described. The application of A-K network models in water resources shows that the method can deal with many problems about classification of no known samples for learning and will have a good applied future.