论文部分内容阅读
竞争学习技术是神经网络用于样本分类中应用十分普通的一种方法。动态竞争学习是一种无导师学习技术，它包括产生网络的方法以及网络中新单元的产生规则，它们都是在学习中动态产生的。该模型可用于复杂数据结构的存储。
Competitive learning technology is a very common method used by neural networks in sample classification. Dynamic competitive learning is a tutorless learning technique that includes methods of generating networks and rules of generating new cells in the network that are both dynamically generated in learning. The model can be used to store complex data structures.