以稀疏子空间聚类以及低秩子空间聚类等基本谱聚类算法为基础，通过
运用核映射算法，融合与数据本身结构相关的局部切线空间函数以及主成分分析
算法建立了可以应对独立子空间聚类、非独立子空间聚类、非线性聚类、混合多
流体聚类问题以及多种含有大数据量的实际问题，包括处理运动分割、人脸识别、
工件识别等情况中的多种类型数据分类的聚类算法，并且引入 Map-Reduce 并行处
理方法优化了算法的计算效率(Based on the basic spectral clustering algorithm such as sparse subspace clustering and low rank subspace clustering, the local tangent space function and the principal component analysis algorithm are established by using the kernel mapping algorithm, which can deal with the independent subspace Clustering, non-independent subspace clustering, nonlinear clustering, mixed multi-fluid clustering problems and a variety of practical problems with large amounts of data, including dealing with motion segmentation, face recognition, workpiece recognition, etc. Data clustering algorithm, and the introduction of Map-Reduce parallel processing method to optimize the computational efficiency of the algorithm)