关键词: 数据降维, 流形学习, 局部线性嵌入, 近似重构系数

Abstract:  Locally Linear Embedding（LLE） is one of the non-linear dimensionality reduction methods which are based on manifold learning. Focused on the existing problem of the selection of the neighborhood and the distribution of sample points, also the time-consuming, an improved LLE algorithm for dimension reduction was proposed. Based on fuzzy clustering theory and distance calculation method, by making use of the characteristic of cluster center including massive information, this paper defined the approximately reconstructing coefficient. The experimental results show that the improved LLE can reduce the influence of the number of neighbors efficiently and obtain good results, also it can reduce time-consuming.


Key words: dimensionality reduction, manifold learning, locally linear embedding, approximate reconstruction coefficient