摘要
现有子空间聚类算法通常假设数据来自多个线性子空间,无法处理时间序列聚类中存在的非线性和时间轴弯曲问题.为了克服这些局限,通过引入核技巧和弹性距离,提出弹性核低秩表示子空间聚类和弹性核最小二乘回归子空间聚类,统称为弹性核子空间聚类,并从理论上证明弹性核最小二乘回归子空间算法的组效应和弹性核低秩表示子空间聚类算法的收敛性.在5个UCR时间序列数据集上的实验表明本文算法的有效性.
In the existing subspace clustering algorithms, it is assumed that the data is derived irom a umon of multiple linear subspace, and these algorithms cannot deal with problems of nonlinear and time warping in time series clustering. To overcome these issues, elastic kernel low rank representation subspace clustering （EKLRR） and elastic kernel least squares regression subspace clustering （EKLSR） are proposed by introducing kernel tricks and elastic distance, and they are called elastic kernel subspace clustering（EKSC）- Moreover, the grouping effect of EKLSR and the convergence of EKLRR are proved theoretically. The experimental results on five UCR datasets show the effectiveness of the proposed algorithms.
出处
《模式识别与人工智能》
EI
CSCD
北大核心
2017年第9期779-790,共12页
Pattern Recognition and Artificial Intelligence
基金
国家自然科学基金项目(No.11571074
71273053)
福建省自然科学基金项目(No.2014J01009)资助~~
关键词
子空间聚类
高斯弹性核
时间轴弯曲
时间序列数据
Subspace Clustering, Gaussian Elastic Kernel, Time Warping, Time Series Data
