论文部分内容阅读
由于差异化分布式数据的稀疏性和混沌性,导致采用粒子群算法进行数据挖掘容易陷入局部收敛,数据挖掘的准确度不好。提出一种基于神经网络和粒子群优化的数据挖掘算法,首先计算差异化分布式数据挖掘的全局核函数和混合核函数,构建挖掘决策模型。以训练误差作为挖掘优化的约束条件,采用粒子群学习的泛化能力进行数据挖掘的聚类中心计算,克服初始聚类中心敏感和易陷入局部最优的缺点。然后采用神经网络模型进行数据聚类处理,实现数据优化挖掘。最后进行仿真实验分析,结果表明,采用本文算法进行差异化分布式数据挖掘具有更高的准确率和更强的收敛能力。
Due to the sparseness and chaos of differentiated distributed data, data mining using Particle Swarm Optimization is prone to be trapped in local convergence and the accuracy of data mining is not good. A data mining algorithm based on neural network and particle swarm optimization is proposed. First, global kernel function and mixed kernel function of differentiated distributed data mining are calculated to build a mining decision model. Taking the training error as the constraint of mining optimization, the clustering center of data mining is calculated by using the generalization ability of particle swarm learning to overcome the shortcomings that the initial cluster centers are sensitive and easy to fall into the local optimum. Then using neural network model for data clustering, data mining optimization. Finally, the simulation experiment is carried out. The results show that the proposed algorithm has better accuracy and convergence ability than the distributed data mining method.