A Cluster Description Method for High Dimensional Data Clustering with Categorical Variables

摘要: High dimensional data clustering is always of great difficulty in research. Before the process accomplished, partition objects unknown. Therefore after process, results final clusters should be presented understandably, which will strictly difficult when it comes to high dimensionality. This paper presents a cluster description schema for with categorical variables. The this uses supremum and infimum represent concisely based on new method given assign non-sample obtained from sample space. distribution requires one-time scan dataset, updates dynamically, can detect isolated objects. Experiments both synthetic real show its effectiveness scalability.