基于几何代数的粒子群协同实用化算法

The Practical Algorithm of Particle Swarm Optimization Based on Geometric Algebra

粒子群协同实用化在web上具有众多的应用,需要聚类大量的代数矩阵才能获得足够的数据,当前研究侧重粒子群协同与矩阵的相关性而没有充分考虑到在几何性之间是具有显著的重叠关系的,这就导致了针对不同的矩阵会具有相同结果,引发重复计算问题。鉴于此,提出了一种基于几何代数的粒子群协同实用化算法,所选择的几何性矩阵具有较高的相关性,而且重叠性较低。实验结果表明,本文的方法能够显著提高粒子群协同实用化的精度和效率。

As particle swarm optimization has been largely applied on the web,a large number of algebraic matrices are required to be clustered to obtain enough data.The current research focuses on the correlation of particle swarm and matric without taking full account of the significant overlapping relationship at the high dimension,which leads to the same result for the different matrices and brings about the repeated computations.Therefore,the paper proposes a particle swarm optimization algorithm based on geometric algebra,for which the selected geometric matrix has a high correlation but low overlapping.The experimental results show that the proposed method can significantly improve the accuracy and efficiency of particle swarm optimization.