第8章聚类分析基本概念和算法知识讲解.pptVIP

聚类分析：基本概念和算法;什么是聚类分析?;什么是一个好的聚类方法?;聚类的复杂性;划分聚类（Partitional Clustering）;层次聚类（Hierarchical Clustering）;互斥的、重叠的、模糊的;不同的簇类型;簇类型: 明显分离的（Well-Separated）;簇类型:基于原型的;簇类型:基于图的;簇类型: 基于密度的（Density-Based）;簇类型: 概念簇（Conceptual Clusters）;K均值聚类;基本K均值算法;数据对象之间的相异度;明可夫斯基距离（Minkowski Distance）;二元数据的相似性度量;余弦相似度;误差平方和（sum of the squared error，SSE）;;选择初始的质心;;二分k均值;二分k均值算法;K means 的优点与缺点;K-means 的局限性;Limitations of K-means: Differing Sizes;Limitations of K-means: Differing Density;Limitations of K-means: Non-globular Shapes;K-means 局限性的克服;Overcoming K-means Limitations;Overcoming K-means Limitations;层次聚类;凝聚的和分裂的层次聚类 ;凝聚的和分裂的层次聚类 ;基本凝聚层次聚类方法;起始步骤 ;中间步骤;中间步骤;最终合并;如何定义cluster间的相似性; ; ; ; ;MIN or 单链 ;分层聚类: MIN;单链的优势;单链的局限性;Cluster Similarity: MAX or 全链;分层聚类: MAX 或全链;Original Points;Original Points;Cluster Similarity: 组平均;分层聚类: Group Average; 优势 对噪音和极端值影响小 局限 偏好球型簇;其它近似度;Cluster Similarity: Ward’s Method;Hierarchical Clustering: Comparison;优点与缺点;DBSCAN算法;;DBSCAN: 核心点, 边界点, 噪声点;DBSCAN 算法;DBSCAN: 选择 EPS and MinPts;;变密度的簇;;;DBSCAN的优点与缺点;簇评估 ;随机数据的聚类结果;Measuring Cluster Validity Via Correlation;Order the similarity matrix with respect to cluster labels and inspect visually. ;Using Similarity Matrix for Cluster Validation;Using Similarity Matrix for Cluster Validation;Using Similarity Matrix for Cluster Validation;Using Similarity Matrix for Cluster Validation;SSE is good for comparing two clusterings or two clusters (average SSE). Can also be used to estimate the number of clusters ;Example Compare SSE of 0.005 against three clusters in random data Histogram shows SSE of three clusters in 500 sets of random data points of size 100 distributed over the range 0.2 – 0.8 for x and y values