论文部分内容阅读
阐述了BP神经网络的原理及学习算法，在结合模糊逻辑推理的基础上提出了一种具有分层结构，能够进行规则自提取、自修正、自学习的复合模糊BP神经网络模型．这种模糊神经网络不仅可以充分利用原有的专家的经验和知识，而且能够从实际数据中通过不断学习获取新的知识和推理规则．同时，在相应的网络权值训练中引入了遗传算法和模糊逻辑控制器的优化求解思想．还进一步探讨了将这种网络模型用于汇率分析系统的形式和方法．
The principle and learning algorithm of BP neural network are expounded. Based on the fuzzy logic inference, a model of complex fuzzy BP neural network with hierarchical structure, capable of rule self-extracting, self-correcting and self-learning is proposed. This fuzzy neural network not only can make full use of the experience and knowledge of the original experts, but also can acquire new knowledge and reasoning rules through continuous learning from the actual data. At the same time, the optimization algorithm of genetic algorithm and fuzzy logic controller is introduced in the corresponding network weight training. The form and method of using this network model for the exchange rate analysis system are further explored.