一种基于子空间聚类的局部相关性可视分析方法

Local Correlation Visual Analysis Based on Subspace Clustering

数据子集局部存在的维度相关性往往被数据集全体所掩盖.为了发现有意义的数据子集,并揭示其表达的维度局部相关性,提出一种局部相关性可视分析方法.首先采用基于测地距离和局部子空间距离的二维散点图揭示子空间聚类模式;然后基于近似覆盖面积和平均距离进行相关显著性估计,给出可能具有局部相关性的二维子空间推荐;最后实现了可视分析系统,并通过案例分析验证了可视分析系统的有效性.

The dimension correlations which exist in subset of data are often obscured in the full dataset. Wepropose a local correlation visual analysis approach to detect meaningful data subset and reveal local dimensioncorrelations. First, a scatter plot is adopted to visually reveal the subspace cluster. The two dimensions of thescatter plot are defined based on geodesic distance and the distance between local subspaces correspondingly.Next, an estimation for correlation significance is proposed based on covering area and mean distance of the data.Subsequently, the 2-dimensional subspaces which reveal local correlations are suggested. Last, a visual analysissystem is implemented and case studies demonstrates the effectiveness and efficiency of our system.