Cutoff for Exact Recovery of Gaussian Mixture Models
Cutoff for Exact Recovery of Gaussian Mixture Models
复制标题
DOI：
10.1109/tit.2021.3063155
复制
发表时间：
2021-06-01
影响因子：
2.5
通讯作者：
Yang, Yun
中科院分区：
文献类型：
--
作者：
Chen, Xiaohui;Yang, Yun
我们确定的信息理论的截断值的聚类中心的分离，在K-分量高斯混合模型具有相等的集群大小的集群标签的准确恢复。此外，我们表明，一个半定规划（SDP）放松的K-均值聚类方法实现精确恢复，而无需假设对称的聚类中心这样尖锐的阈值。
We determine the information-theoretic cutoff value on separation of cluster centers for exact recovery of cluster labels in a K-component Gaussian mixture model with equal cluster sizes. Moreover, we show that a semidefinite programming (SDP) relaxation of the K-means clustering method achieves such sharp threshold for exact recovery without assuming the symmetry of cluster centers.