Clustering Objects Generated by Linear Regression Models
提出一种针对线性模型不同状态下生成对象的聚类方法，利用奇异值分解计算质心，并通过两种启发式迭代算法求解组合优化问题，仿真验证了有效性。
Abstract This article describes a new clustering method designed for objects generated by a linear model with different states. The objective is to divide the objects into clusters, with each cluster containing only the objects taken when the system is in one particular state. The cluster centroids are matrices of predetermined rank and can be computed by the singular value decomposition (SVD) algorithm. The clustering problem is formulated as a combinatorial optimization problem. Two heuristic iterative algorithms that ensure the decrease of the objective function are then proposed. The simulation examples are given to show the effectiveness of the methodology.