分群理論的相關名詞.doc

Clustering: 共有個要點 算出某點與其餘之間的距離, 選出和自己最近的那點. 依此類推 兩點之間的距離如何算出? 利用Euclidean distance Euclidean distance?between points??and?, in?Euclidean?n-space, is defined as: 群集 (clustering)分為3種方法: (1)分割表示法-partitional method (2)階層表示法hierarchical method (3)密集度導向表示法 density-based method (1)分割式分群(partitional method): 包括k-means, k-medoids method, ISODATA k-means-以質量中心為群集之中心點k-medoids method—以每一群集中的中心點當為群集中心 ISODATA(2)階層式分群(hierarchical method): 包括AGNES, DIANA(3)分割表示法-partitional method: 包括k-means, k-medoids method, ISODATA 階層分級法(hierarchical clustering)的圖示法: A?dendrogram?(from?Greek?dendron?tree,?-gramma?drawing) is a?tree?diagram frequently used to illustrate the arrangement of the clusters produced by a?clustering algorithm. Dendrograms are often used in?computational biology?to illustrate the clustering of?genes. For a clustering example, suppose this data is to be clustered using?Euclidean distance?as the?distance metric. Raw data The?hierarchical clustering?dendrogram would be as such: Traditional representation 群集之間的的相似程度:有 single-linkage(以兩個群集點和點之間最近的距離計算) complete-linkage(以兩個群集點和點之間最遠的距離計算)mean-linkage(以兩個群集點和點之間距離的平均值計算, 如UPGMA) Ward’s criterion(以兩個群集點合併後的方差[variance]增加的量來計算)