论文部分内容阅读
研究粒子群K均值聚类算法问题,针对传统粒子群K均值算法容易陷入局部最优解,出现早熟收敛的缺点,提出一种基于云模型改进的粒子群K均值聚类算法,使用X条件云发生器自适应地调整粒子个体惯性权重的方法,保证惯性权重会逐渐减小而又不失随机性。根据个体适应度的优劣将粒子群分为三个子群,在每次迭代时都保证仍有一个子群的粒子在进行全局搜索,避免算法陷入局部最优和早熟收敛。在典型数据集上的仿真结果表明,改进算法相比其他聚类算法得到较好的聚类准确率和较快的收敛速度,是一种行之有效的方法。
In order to solve the problem that the traditional K-means clustering algorithm is easy to fall into the local optimal solution and has premature convergence, a K-means clustering algorithm based on improved cloud model is proposed. Using the X-conditional cloud The generator adaptively adjusts the inertia weight of individual particles to ensure that the inertia weight will be gradually reduced without losing randomness. Particle swarm is divided into three subgroups according to the fitness of individual fitness. At each iteration, it is guaranteed that there is still a subgroup of particles in the global search, to avoid the algorithm into the local optimum and premature convergence. The simulation results on the typical data set show that the improved algorithm has better clustering accuracy and faster convergence rate than other clustering algorithms, which is an effective method.