基于聚类的模糊神经网络预测系统

CLUSTERING -BASED FUZZY NEURAL NETWORK PREDICTOR

本文在Ｔａｋａｇｉ提出的神经网络驱动的模糊推理基础上，提出了基于聚类的新模糊神经网络结构和学习算法．该模糊神经网络能够利用聚类信息大大减少网络的学习量，并使得系统的应用性能有所提高．该结构和算法非常适合于大型复杂系统的自适应建模和数据开采等领域．

The paper has presented a new fuzzy neural network and its learning algorithm based on clustering. The structure is built on Takagi's NN driven fuzzy reasoning. The FNN could reduce learning workload evendently as well as promote system applicable performance and abtain higher adaption. The structure and algorihm are very fittable to large scale and complex system modeling and data mining. In the end ,the simulation result is given.