基于信息熵的混合属性数据谱聚类算法

Entropy-based spectral clustering algorithm for mixed type data

针对传统的聚类算法只能处理单属性的数据,不能很好地处理混合属性数据的聚类问题,以及目前大多数混合属性数据聚类算法对初始化敏感,不能处理任意形状的数据的问题,提出一种基于信息熵的混合属性数据谱聚类算法,用于处理混合类型数据。提出了一种新的相似性度量方式,利用谱聚类算法中的数值型数据构成的高斯核函数矩阵与新的基于信息熵的分类型数据构成的影响因子矩阵相结合代替了传统的相似度矩阵,新的相似度矩阵避免了数值属性与分类属性数据之间的转换和参数调整;把新的相似度矩阵运用到谱聚类算法中,以便于处理任意形状的数据,最终得出聚类结果。通过在UCI的数据集上的实验表明,该算法能有效地处理混合属性数据的聚类问题,且具有较高的稳定性以及良好的鲁棒性。

The problem that the traditional clustering algorithm can only deal with single attribute data and cannot handle the clustering problem of mixed type data very well. Most of the clustering algorithms for mixed type data currently have the problem of initializing sensitive and cannot handle the data of arbitrary shape. This paper proposed an entropy-based spectral clustering algorithm for mixed type data to deal with mixed type data. First, it proposed a new similarity measure, it used the numerical data in the spectral clustering algorithm to constitute a Gaussian kernel function of the matrix, and used the classification data to constitute an entropy-based the influence factor of the matrix. A new similarity matrix combined these two matrices. Instead of the traditional similarity matrix, it proposed the new similarity matrix avoid feature transformation and parameter adjustment between the numerical data and the classification data. Then, it applied the new similarity matrix to the spectral clustering algorithm so as to deal with the data of arbitrary shape, and finally got the clustering result. Experiments on UCI data sets show that this algorithm can effectively deal with the clustering problem of mixed attribute data, with high stability and good robustness.