论文部分内容阅读
为解决烟草配送中配送区域划分问题,提出了一种改进的K-means聚类算法。计算每个点的密度并取其中最大的K个点作为初始聚类中心;通过分析比较边缘点到聚类中心距离与所有点间的平均距离,在可选范围内优先考虑边缘点,以避免边缘点对整体最优性的干扰。实例分析表明,该算法有较好的全局收敛性,有效地克服了传统K-means算法收敛于局部最优点和忽视边缘点重要性的缺点。
In order to solve the problem of distribution area distribution in tobacco distribution, an improved K-means clustering algorithm is proposed. Calculate the density of each point and take the largest K points as the initial cluster centers. By analyzing and comparing the distance between the edge points and the cluster center and the average distance between all points, give priority to the edge points in the optional range to avoid Interference of edge point to global optimality. The case study shows that this algorithm has good global convergence and overcomes the shortcomings of the traditional K-means algorithm converging to the local optimum and ignoring the importance of the edge.