用于构建数据概要，用于分类、模式识别、假设生成和测试；用于异常检测，检测远离群簇的点。(Used to build data outlines for classification, pattern recognition, hypothesis generation and testing; for anomaly detection, detecting points away from the cluster.)