WJ-CH10-模式识别-聚类算法-02 英文版 教学教案.ppt

王杰(博士/教授/博导) 郑州大学电气工程学院 0371 wj@zzu.edu.cn 模式识别 Pattern Recognition Chapter 10(II) CLUSTERING ALGORITHMS CLUSTERING ALGORITHMS Number of possible clusterings Let X={x1,x2,…,xN}. Question: In how many ways the N points can be assigned into m groups? Answer: Examples: A way out: Consider only a small fraction of clusterings of X and select a “sensible” clustering among them Question 1: Which fraction of clusterings is considered? Question 2: What “sensible” means? The answer depends on the specific clustering algorithm and the specific criteria to be adopted. MAJOR CATEGORIES OF CLUSTERING ALGORITHMS Sequential: A single clustering is produced. One or few sequential passes on the data. Hierarchical: A sequence of (nested) clusterings is produced. Agglomerative Matrix theory Graph theory Divisive Combinations of the above (e.g., the Chameleon algorithm.) Cost function optimization. For most of the cases a  single clustering is obtained. Hard clustering (each point belongs exclusively to a single cluster): Basic hard clustering algorithms (e.g., k-means) k-medoids algorithms Mixture decomposition Branch and bound Simulated annealing Deterministic annealing Boundary detection Mode seeking Genetic clustering algorithms Fuzzy clustering (each point belongs to more than one clusters simultaneously). Possibilistic clustering (it is based on the possibility of a point to belong to a cluster). Basic Sequential Clustering Algorithm (BSAS) m=1 \{number of clusters}\ Cm={x1} For i=2 to N Find Ck: d(xi,Ck)=min1?j?md(xi,Cj) If (d(xi,Ck)Θ) AND (mq) then m=m+1 Cm={xi} Else Ck=Ck?{xi} Where necessary, update representatives (*) End {if} End {for} (*) When the mean vector mC is used as representative of the cluster C with nc elements, the updating in the light of a new vector x becomes mCnew=(nC mC + x) / (nC+1) Remarks: The order of presentation of the data in the algorithm plays important role in the clustering results. Different