Upcoming SlideShare
×

# Dbm630 lecture09

1,546 views

Published on

Published in: Technology, Education
1 Like
Statistics
Notes
• Full Name
Comment goes here.

Are you sure you want to Yes No
• Be the first to comment

Views
Total views
1,546
On SlideShare
0
From Embeds
0
Number of Embeds
2
Actions
Shares
0
97
0
Likes
1
Embeds 0
No embeds

No notes for slide

### Dbm630 lecture09

1. 1. DBM630: Data Mining and Data Warehousing MS.IT. Rangsit University Semester 2/2011 Lecture 9 Clustering by Kritsada Sriphaew (sriphaew.k AT gmail.com)1
2. 2. Topics What is Cluster Analysis? Types of Attributes in Cluster Analysis Major Clustering Approaches  Partitioning Algorithms  Hierarchical Algorithms 2 Data Warehousing and Data Mining by Kritsada Sriphaew
3. 3. Classification vs. Clustering Classification: Supervised learning Learns a method for predicting the instance class from pre-labeled (classified) instances3 Clustering Analysis
4. 4. Clustering Unsupervised learning: Finds “natural” grouping of instances given un-labeled data4 Clustering Analysis
5. 5. Clustering Methods Many different method and algorithms:  For numeric and/or symbolic data  Deterministic vs. probabilistic  Exclusive vs. overlapping  Hierarchical vs. flat  Top-down vs. bottom-up5 Clustering Analysis
6. 6. What is Cluster Analysis ? Cluster: a collection of data objects  High similarity of objects within a cluster  Low similarity of objects across clusters Cluster analysis  Grouping a set of data objects into clusters Clustering is an unsupervised classification: no predefined classes Typical applications  As a stand-alone tool to get insight into data distribution  As a preprocessing step for other algorithms 6 Clustering Analysis
7. 7. General Applications of Clustering Pattern Recognition  In biology, derive plant and animal taxonomies, categorize genes Spatial Data Analysis  create thematic maps in GIS by clustering feature spaces  detect spatial clusters and explain them in spatial data mining Image Processing Economic Science (e.g. market research)  discovering distinct groups in their customer bases and characterize customer groups based on purchasing patterns WWW  Document classification  Cluster Weblog data to discover groups of similar access patterns 7 Clustering Analysis
8. 8. Examples of Clustering Applications Marketing: Help marketers discover distinct groups in their customer bases, and then use this knowledge to develop targeted marketing programs Land use: Identification of areas of similar land use in an earth observation database Insurance: Identifying groups of motor insurance policy holders with a high average claim cost City-planning: Identifying groups of houses according to their house type, value, and geographical location Earth-quake studies: Observed earth quake epicenters should be clustered along continent faults 8 Clustering Analysis
9. 9. Clustering Evaluation Manual inspection Benchmarking on existing labels Cluster quality measures  distance measures  high similarity within a cluster, low across clusters 9 Clustering Analysis
10. 10. Criteria for Clustering A good clustering method will produce high quality clusters with  high intra-class similarity  low inter-class similarity The quality of a clustering result depends on:  both the similarity measure used by the method and its implementation to the new cases  ability to discover some or all of the hidden patterns 10 Clustering Analysis
11. 11. Requirements of Clustering in Data Mining Scalability Ability to deal with different types of attributes Discovery of clusters with arbitrary shape Minimal requirements for domain knowledge to determine input parameters Able to deal with noise and outliers Insensitive to order of input records High dimensionality Incorporation of user-specified constraints Interpretability and usability 11 Clustering Analysis
12. 12. The distance function Simplest case: one numeric attribute A  Distance(X,Y) = A(X) – A(Y) Several numeric attributes:  Distance(X,Y) = Euclidean distance between X,Y Nominal attributes: distance is set to 1 if values are different, 0 if they are equal Are all attributes equally important?  Weighting the attributes might be necessary 12 Clustering Analysis
13. 13. From Data Matrix to Similarity or Dissimilarity Matrices Data matrix (or object-by-attribute structure)  m objects with n attributes, e.g., relational data  x11  x1 j  x1n         x  xij  xin   i1          xm1   xmj  xmn   Similarity and dissimilarity matrices  a collection of proximities for all pairs of m objects. 1   0          s  1  d  0   i1   i1        sij  s ji       dij  d ji 13  m1 s  smj  1 0  sij  1  d m1   d mj  0 0  dijAnalysis  Clustering 1
14. 14. Distance Functions (Overview) To transform a data matrix to similarity or dissimilarity matrices, we need a definition of distance. Some definitions of distance functions depend on the type of attributes  interval-scaled attributes  Boolean attributes  nominal, ordinal and ratio attributes. Weights should be associated with different attributes based on applications and data semantics. It is hard to define “similar enough” or “good enough”  the answer is typically highly subjective. 14 Clustering Analysis
15. 15. Similarity and Dissimilarity Between Objects (I) Distances are normally used to measure the similarity or dissimilarity between two data objects Some popular ones include: Minkowski distance: dij  q (| xi1  x j1 |q  | xi 2  x j 2 |q ... | xip  x jp |q ) where i = (xi1, xi2, …, xip) and j = (xj1, xj2, …, xjp) are two p-dimensional data objects, and q is a positive integer If q = 1, d is Manhattan distance dij  | xi1  x j1 |  | xi 2  x j 2 | ... | xip  x jp | 15 Clustering Analysis
16. 16. Similarity and Dissimilarity Between Objects (II) If q = 2, d is Euclidean distance: dij  (| xi1  x j1 |2  | xi 2  x j 2 |2 ... | xip  x jp |2 ) Properties d(i,j) >= 0 d(i,i) = 0 d(i,j) = d(j,i) d(i,j) <= d(i,k) + d(k,j) Also one can use weighted distance, parametric Pearson product moment correlation, or other dissimilarity measures. d  (w | x  x | w | x  x | ... w | x  x | ) ij 1 i1 j1 2 2 i2 j2 2 i ip jp 2 16 Clustering Analysis
17. 17. Types of Attributes in Clustering Interval-scaled attributes  Continuous measures of a roughly linear scale Binary attributes  Two-state measures: 0 or 1 Nominal, ordinal, and ratio attributes  More than two states, nominal or ordinal or nonlinear scale Mixed types  Mixture of interval-scaled, symmetric binary, asymmetric binary, nominal, ordinal, or ratio-scaled attributes 17 Clustering Analysis: Types of Attributes
18. 18. Interval-valued Attributes Standardize data  Calculate the mean absolute deviation: s f  1 (| x1 f  m f |  | x2 f  m f | ... | xnf  m f |) n where m f  1 (x1 f  x2 f n  ...  xnf ) .  Calculate the standardized measurement (mean-absolute-based z- xif  m f score) zif  sf Using mean absolute deviation is more robust than using standard deviation since the z-scores of outliers do not become too small. Hence, the outliers remain detectable. 18 Clustering Analysis: Types of Attributes
19. 19. A binary variable contains two possible outcomes: 1Binary Attributes (positive/present) or 0 (negative/absent). • If there is no preference for A contingency table for binary data which outcome should be coded as 0 and which as 1, the binary variable is Object j called symmetric. 1 0 sum • If the outcomes of a binary variable are not equally 1 a b a b important, the binary variable is called asymmetric, such as "is Object i 0 c d cd color-blind" for a human being. The most important outcome sum a  c b  d p is usually coded as 1 (present) and the other is coded as 0 (absent). Simple matching coefficient (invariant, if the binary variable is symmetric): bc dij  abcd Jaccard coefficient (noninvariant if the binary variable is asymmetric): bc dij  19 a  b  cAnalysis:Types of Attributes Clustering
20. 20. Dissimilarity on Binary Attributes Example Name Gender Fever Cough Test-1 Test-2 Test-3 Test-4 Jack M Y N P N N N Mary F Y N P N P N Jim M Y P N N N N The gender is symmetric attribute and the remaining attributes are asymmetric binary. Here, let the values Y and P be set to 1, and the value N be set to 0. Then calculate only asymmetric binary 0 1 d ( jack , mary)   0.33 2  0 1 11 d ( jack , jim )   0.67 111 1 2 d ( jim , mary)   0.75 11 2 20 Clustering Analysis: Types of Attributes
21. 21. Nominal Attributes A generalization of the binary variable in that it can take more than 2 states, e.g., red, yellow, blue, green Method 1: Simple matching  m is the # of matches, p is the total # of nominal attributes pm dij  p Method 2: Use a large number of binary variables  creating a new binary variable for each of the M nominal states 21 Clustering Analysis: Types of Attributes
22. 22. Ordinal Attributes An ordinal variable can be discrete or continuous order is important, e.g., rank Can be treated like interval-scaled replacing xif by their rank rif { ,..., M f } 1 map the range of each variable onto [0, 1] by replacing i-th object in the f-th variable by rif 1 zif  M f 1 compute the dissimilarity using methods for interval-scaled variables 22 Clustering Analysis: Types of Attributes
23. 23. Ratio-Scaled Attributes Ratio-scaled variable: a positive measurement on a nonlinear scale, approximately at exponential scale, such as AeBt or Ae-Bt Methods:(1) treat them like interval-scaled attributes not a good choice!(2) apply logarithmic transformation yif = log(xif)(3) treat them as continuous ordinal data and treat their rank as interval-scaled. 23 Clustering Analysis: Types of Attributes
24. 24. Mixed Types A database may contain all the six types of attributes  symmetric binary, asymmetric binary, nominal, ordinal, interval and ratio. One may use a weighted formula to combine their effects.  p 1 ij f ) dij f ) ( ( dij  f p  f 1 ij f ) (  f is binary or nominal:  dij(f) = 0 if xif = xjf , or dij(f) = 1 otherwise  f is interval-based: use the normalized distance  f is ordinal or ratio-scaled  compute ranks rif and  treat (normalized) zif as interval-scaled 24 Clustering Analysis: Types of Attributes
25. 25. Major Clustering Approaches Partitioning algorithms: Construct various partitions and then evaluate them by some criterion Hierarchy algorithms: Create a hierarchical decomposition of the set of data (or objects) using some criterion Density-based: based on connectivity and density functions Grid-based: based on a multiple-level granularity structure Model-based: A model is hypothesized for each of the clusters and the idea is to find the best fit of that model to each other 25 Clustering Analysis: Clustering Approaches
26. 26. Partitioning Approach Construct a partition of a database D of n objects into a set of k clusters Given a k, find a partition of k clusters that optimizes the chosen partitioning criterion Global optimal: exhaustively enumerate all partitions Heuristic methods: k-means and k-medoids algorithms k-means (MacQueen’67): Each cluster is represented by the center of the cluster k-medoids or PAM (Partition around medoids) (Kaufman & Rousseeuw’87): Each cluster is represented by one of the objects in the cluster 26 Clustering Analysis: Partitioning Algorithms
27. 27. The K-Means Clustering Method(Overview) Given k, the k-means algorithm is implemented in 4 steps:  Partition objects into k nonempty subsets  Compute seed points as the centroids of the clusters of the current partition. The centroid is the center (mean point) of the cluster.  Assign each object to the cluster with the nearest seed point.  Go back to Step 2, stop when no more new assignment. 27 Clustering Analysis: Partitioning Algorithms
28. 28. The K-Means Clustering Method(An Graphical Example) 10 10 9 9 8 8 7 7 6 6 5 5 4 4 3 3 2 2 1 1 0 0 0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10 10 10 9 9 8 8 7 7 6 6 5 5 4 4 3 3 2 2 1 1 0 0 0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 1028 Clustering Analysis: Partitioning Algorithms
29. 29. K-means example, step 1Given k=3, k1 YPick 3 k2initialclustercenters(randomly) k3 X 29 Clustering Analysis: Partitioning Algorithms
30. 30. K-means example, step 2 k1 Y k2Assigneach pointto the closestclustercenter k3 X 30 Clustering Analysis: Partitioning Algorithms
31. 31. K-means example, step 3 k1 k1 YMove k2each clustercenter k3 k2to the meanof each cluster k3 X 31 Clustering Analysis: Partitioning Algorithms
32. 32. K-means example, step 4Reassign k1points Yclosest to adifferent newcluster center k3Q: Which k2points arereassigned? X 32 Clustering Analysis: Partitioning Algorithms
33. 33. K-means example, step 4 … k1 YA: threepoints withanimation k3 k2 X 33 Clustering Analysis: Partitioning Algorithms
34. 34. K-means example, step 4b k1 Yre-computeclustermeans k3 k2 X 34 Clustering Analysis: Partitioning Algorithms
35. 35. K-means example, step 5 k1 Y k2move clustercenters to k3cluster means X 35 Clustering Analysis: Partitioning Algorithms
36. 36. Problems to be considered What can be the problems with K-means clustering? Result can vary significantly depending on initial choice of seeds (number and position) Can get trapped in local minimum initial cluster  Example: centers instances Q: What can be done? A: To increase chance of finding global optimum: restart with different random seeds. What can be done about outliers? 36 Clustering Analysis: Partitioning Algorithms
37. 37. The K-Means Clustering Method(Strength and Weakness) Strength  Relatively efficient: O(tkn), where n is # objects, k is # clusters, and t is # iterations. Normally, k, t << n  Good for finding clusters with spherical shapes  Often terminates at a local optimum. The global optimum may be found using techniques such as: deterministic annealing and genetic algorithms Weakness  Applicable only when mean is defined, then what about categorical data?  Need to specify k, no. of clusters, in advance  Unable to handle noisy data and outliers  Not suitable to discover clusters with non-convex shapes 37 Clustering Analysis: Partitioning Algorithms
38. 38. The K-Means Clustering Method(Variations – I) A few variants of the k-means which differ in  Selection of the initial k means Mean of 1, 3, 5, 7, 9 is 5 Mean of 1, 3, 5, 7, 1009 is 205  Dissimilarity calculations Median of 1, 3, 5, 7, 1009 is 5 Median advantage: not affected  Strategies to calculate cluster means by extreme values  K-medoids – instead of mean, use medians of each cluster  For large databases, use sampling Handling categorical data: k-modes (Huang’98)  Replacing means of clusters with modes  Using new dissimilarity measures to deal with categorical objects  Using a frequency-based method to update modes of clusters  A mixture of categorical/numerical data: k-prototype method 38 Clustering Analysis: Partitioning Algorithms
39. 39. The K-Medoids Clustering Method(Overview) Find representative objects, called medoids, in clusters PAM (Partitioning Around Medoids, 1987)  starts from an initial set of medoids and iteratively replaces one of the medoids by one of the non-medoids if it improves the total distance of the resulting clustering  PAM works effectively for small data sets, but does not scale well for large data sets CLARA (Kaufmann & Rousseeuw, 1990) CLARANS (Ng & Han, 1994): Randomized sampling 39 Clustering Analysis: Partitioning Algorithms
40. 40. The K-Medoids Clustering Method(PAM - Partitioning Around Medoids) PAM (Kaufman and Rousseeuw, 1987), built in Splus Use real object to represent the cluster  Select k representative objects arbitrarily  For each pair of non-selected object h and selected object i, calculate the total swapping cost Sih  For each pair of i and h,  If Sih < 0, i is replaced by h  Then assign each non-selected object to the most similar representative object  repeat steps 2-3 until there is no change 40 Clustering Analysis: Partitioning Algorithms
41. 41. PAM example Cluster the following data set of ten objects into two clusters i.e k = 2. X1 2 6 X2 3 4 X3 3 8 X4 4 7 X5 6 2 X6 6 4 X7 7 3 X8 7 4 X9 8 5 X10 7 641 Clustering Analysis
42. 42. PAM example, step 1  Initialise k medoids. Let assume c1 = (3,4) and c2 = (7,4)  Calculating distance so as to associate each data object to its nearestCost medoid. Assume that cost is Cost calculated )using Minkowski distance metric with r =(distan c 1 Data objects (X (distan 2 c Data objects (X ) 1. i ce) i ce)3 4 2 6 3 7 4 2 6 73 4 3 8 4 7 4 3 8 83 4 4 7 4 7 4 4 7 63 4 6 2 5 7 4 6 2 33 4 6 4 3 7 4 6 4 13 4 7 3 5 7 4 7 3 13 4 8 5 6 7 4 8 5 2 42 Clustering Analysis3 4 7 6 6 7 4 7 6 2
43. 43. PAM example, step 1b  Then the clusters become: Cluster1 = {(3,4)(2,6)(3,8)(4,7)} Cluster2 = {(7,4)(6,2)(6,4)(7,3)(8,5)(7,6)}  Total cost = 3 + 4 +Cost+ 3 + 1 + 1 + 2 + 2 = 20 4 Cost Data objects Data objects c1 (distan c2 (distan (Xi) (Xi) ce) ce)3 4 2 6 3 7 4 2 6 73 4 3 8 4 7 4 3 8 83 4 4 7 4 7 4 4 7 63 4 6 2 5 7 4 6 2 33 4 6 4 3 7 4 6 4 13 4 7 3 5 7 4 7 3 13 4 8 5 6 7 4 8 5 2 43 Clustering Analysis3 4 7 6 6 7 4 7 6 2
44. 44. PAM example, step 2 Selection of nonmedoid O′ randomly. Let us  assume O′ = (7,3). So now the medoids are c1(3,4) and O′(7,3)  Calculate the cost of new medoid by using the Cost Cost Data objects Data objects formula in )the step1. Total cost = c 1 (X (distan O′ (X ) (distan i ce) i ce)3 3+4+4+2+2+1+3+33 = 22 7 4 2 6 3 2 6 83 4 3 8 4 7 3 3 8 93 4 4 7 4 7 3 4 7 73 4 6 2 5 7 3 6 2 23 4 6 4 3 7 3 6 4 23 4 7 3 5 7 3 7 4 13 4 8 5 6 7 3 8 5 3 44 Clustering Analysis3 4 7 6 6 7 3 7 6 3
45. 45. PAM example, step 2b So cost of swapping medoid from c2 to O′ is S = current total cost – past total cost = 22-20= 2 >0 So moving to O′ would be bad idea, so the previous choice was good, and algorithm terminates here (i.e there is no change in the medoids). It may happen some data points may shift from one cluster to another cluster depending upon their closeness to medoid 45 Clustering Analysis
46. 46. CLARA (Clustering Large Applications) (1990) CLARA (Kaufmann and Rousseeuw in 1990)  Built in statistical analysis packages, such as S+ It draws multiple samples of the data set, applies PAM on each sample, and gives the best clustering as the output Strength: deals with larger data sets than PAM Weakness:  Efficiency depends on the sample size  A good clustering based on samples will not necessarily represent a good clustering of the whole data set if the sample is biased 46 Clustering Analysis: Partitioning Algorithms
47. 47. CLARANS (“Randomized” CLARA) (1994) CLARANS (A Clustering Algorithm based on Randomized Search) (Ng and Han’94) CLARANS draws sample of neighbors dynamically The clustering process can be presented as searching a graph where every node is a potential solution, that is, a set of k medoids If the local optimum is found, CLARANS starts with new randomly selected node in search for a new local optimum It is more efficient and scalable than both PAM and CLARA Focusing techniques and spatial access structures may further improve its performance (Ester et al.’95) 47 Clustering Analysis: Partitioning Algorithms
48. 48. The Partition-Based Clustering(Discussion) Result can vary significantly based on initial choice of seeds Algorithm can get trapped in a local minimum  Example: four instances at the vertices of a two- dimensional rectangle  Local minimum: two cluster centers at the midpoints of the rectangle’s long sides Simple way to increase chance of finding a global optimum: restart with different random seeds 48 Clustering Analysis: Hierarchical Algorithms
49. 49. Hierarchical Clustering Use distance matrix as clustering criteria. This method does not require the number of clusters k as an input, but needs a termination condition Step 0 Step 1 Step 2 Step 3 Step 4 Agglomerative (AGNES) a ab b abcde c cde d de e Divisive Step 4 Step 3 Step 2 Step 1 Step 0 (DIANA) 49 Clustering Analysis: Hierarchical Algorithms
50. 50. AGNES (Agglomerative Nesting) Introduced in Kaufmann and Rousseeuw (1990) Implemented in statistical analysis packages, e.g., Splus Use the Single-Link method and the dissimilarity matrix. Merge nodes that have the least dissimilarity Go on in a non-descending fashion Eventually all nodes belong to the same cluster 10 10 10 9 9 9 8 8 8 7 7 7 6 6 6 5 5 5 4 4 4 3 3 3 2 2 2 1 1 1 0 0 0 0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10 50 Clustering Analysis: Hierarchical Algorithms
51. 51. Dendrogram for Hierarchical Clustering Decompose data objects into a several levels of nested partitioning (tree of clusters), called a dendrogram. A clustering of the data objects is obtained by cutting the dendrogram at the desired level, then each connected component forms a cluster. 51 Clustering Analysis: Hierarchical Algorithms
52. 52. DIANA - Divisive Analysis Introduced in Kaufmann and Rousseeuw (1990) Implemented in statistical analysis packages, e.g., Splus Inverse order of AGNES Eventually each node forms a cluster on its own 10 10 10 9 9 9 8 8 8 7 7 7 6 6 6 5 5 5 4 4 4 3 3 3 2 2 2 1 1 1 0 0 0 0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10 52 Clustering Analysis: Hierarchical Algorithms
53. 53. Hierarchical Clustering Major weakness of agglomerative clustering methods  do not scale well: time complexity of at least O(n2), where n is the number of total objects  can never undo what was done previously53 Clustering Analysis: Hierarchical Algorithms
54. 54. Distances - Hierarchical Clustering(Overview) Four measures for distance between clusters are:  Single linkage (Minimum distance): dmin (Ci , C j )  minpCi , pCj p  p  Complete linkage (Maximum distance): dmax (Ci , C j )  max pCi , pCj p  p  Centroid comparison (Mean distance): d mean (Ci , C j )  mi  m j  Element comparison (Average distance): 1 d avg (Ci , C j )    p  p 54 ni n j pCi pC j Analysis: Hierarchical Algorithms Clustering
55. 55. Distances - Hierarchical Clustering (Graphical Representation)  Four measures for distance between clusters are (1) single linkage, (2) complete linkage, (3) centroid comparison and (4) element comparison Cluster 1 (1)single Cluster 2 (4) Element comparison: x x average distance among all elements in two clusters (3)centroid(2)complete Cluster 3 x x = centroids 55 Clustering Analysis: Hierarchical Algorithms
56. 56. Practice Use single and complete link agglomerative clustering to group the data described by the following distance matrix. Show the dendrograms. A B C D A 0 1 4 5 B 0 2 6 C 0 3 D 056 Clustering Analysis
57. 57. Other Clustering Methods Incremental Clustering Probability-based Clustering, Bayesian Clustering EM AlgorithmCluster Schemes Partitioning Methods Hierarchical Methods Density-Based Methods Grid-Based Methods Model-Based Clustering Methods 57 Clustering Analysis
58. 58. Advanced Method: BIRCH (Overview) Balanced Iterative Reducing and Clustering using Hierarchies [Tian Zhang, Raghu Ramakrishnan, Miron Livny, 1996] Incremental, hierarchical, one scan Save clustering information in a tree Each entry in the tree contains information about one cluster New nodes inserted in closest entry in tree Only works with "metric" attributes  Must have Euclidean coordinates Designed for very large data sets  Time and memory constraints are explicit  Treats dense regions of data points as sub-clusters  Not all data points are important for clustering  Only one scan of data is necessary 58 Clustering Analysis
59. 59. BIRCH (Merits) Incremental, distance-based approach Decisions are made without scanning all data points, or all currently existing clusters Does not need the whole data set in advance Unique approach: Distance-based algorithms generally need all the data points to work Make best use of available memory while minimizing I/O costs Does not assume that the probability distributions on attributes is independent 59 Clustering Analysis
60. 60. BIRCH – Clustering Feature and Clustering Feature Tree BIRTH introduces two concepts, clustering feature and clustering feature tree (CF Tree), which are used to summarize cluster representations. These structures help the clustering method achieve good speed and scalability in large databases and make it effective for incremental and dynamic clustering of incoming object Given n d-dimensional data objects or points in a cluster, we can define the centroid x0, radius R and diameter D of the cluster as follows: 60 Clustering Analysis
61. 61. BIRCH – Centroid, Radius and Diameter• Given a cluster of instances , we define: • Centroid: the center of a cluster • Radius: average distance from member points to centroid • Diameter: average pair-wise distance within a cluster 61 Clustering Analysis
62. 62. BIRCH – Centroid Euclidean and Manhattan distances• The centroid Euclidean distance and centroid Manhattan distance are defined between any two clusters. • Centroid Euclidean distance • Centroid Manhattan distance 62 Clustering Analysis
63. 63. BIRCH(Average inter-cluster, Average intra-cluster, Variance increase)• The average inter-cluster, the average intra-cluster, and the variance increase distances are defined as follows • Average inter-cluster • Average intra-cluster • Variance increase distances 63 Clustering Analysis
64. 64. Clustering Feature CF = (N,LS,SS)  N: Number of points in cluster  LS: Sum of points in the cluster  SS: Sum of squares of points in the cluster CF Tree  Balanced search tree  Node has CF triple for each child  Leaf node represents cluster and has CF value for each subcluster in it.  Subcluster has maximum diameter 64 Clustering Analysis
65. 65. Clustering Feature Vector (3,4) (4,7) Clustering Feature: CF = (N, LS, SS) (2,6) (3,8) (4,5) N: Number of data points LS: Ni=1=Xi 10 CF = (5, (16,30),(54,190)) SS: Ni=1=Xi2 9 8 7 6 (6,2) (8,4) (7,2) (8,5) 5 4 3 2 (7,4) 1 0 0 1 2 3 4 5 6 7 8 9 10 CF = (5, (36,17),(262,65))CF = (10, (52,47),(316,255)) CF1  CF2  ( N1  N 2 , LS1  LS2 , SS1  SS2 ) 65 Clustering Analysis
66. 66. Merging two clustersCluster {Xi}: Cluster {Xj}:i = 1, 2, …, N1 j = N1+1, N1+2, …, N1+N2 Cluster Xl = {Xi} + {Xj}: l = 1, 2, …, N1, N1+1, N1+2, …, N1+N266 Clustering Analysis
67. 67. CF Tree RootB=7 CF1 CF2 CF3 CF6 child1 child2 child3 child6L=6 Non-leaf node CF1 CF2 CF3 CF5 child1 child2 child3 child5 Leaf node Leaf nodeprev CF1 CF2 CF6 next prev CF1 CF2 CF4 next 67 Clustering Analysis
68. 68. Properties of CF-Tree  Each non-leaf node has at most B entries  Each leaf node has at most L CF entries which each satisfy threshold T  Node size is determined by dimensionality of data space and input parameter P (page size)Branching FactorandThread hold 68 Clustering Analysis
69. 69. BIRCH Algorithm (CF-Tree Insertion) Recurse down from root, find the appropriate leaf Follow the "closest"-CF path, w.r.t. D0 / … / D4 Modify the leaf If the closest-CF leaf cannot absorb, make a new CF entry. If there is no room for new leaf, split the parent node Traverse back & up69 Updating CFs on the path Analysis Clustering or splitting nodes
70. 70. Improve Clusters70 Clustering Analysis
71. 71. BIRCH Algorithm (Overall steps)71 Clustering Analysis
72. 72. Details of Each Step Phase 1: Load data into memory  Build an initial in-memory CF-tree with the data (one scan)  Subsequent phases become fast, accurate, less order sensitive Phase 2: Condense data  Rebuild the CF-tree with a larger T  Condensing is optional Phase 3: Global clustering  Use existing clustering algorithm on CF entries  Helps fix problem where natural clusters span nodes Phase 4: Cluster refining  Do additional passes over the dataset & reassign data points to the closest centroid from phase 3  Refining is optional 72 Clustering Analysis
73. 73. Summary of BIRCH BIRCH works with very large data sets Explicitly bounded by computational resources. The computation complexity is O(n), where n is the number of objects to be clustered. Runs with specified amount of memory (P) Superior to CLARANS and k-MEANS Quality, speed, stability and scalability 73 Clustering Analysis
74. 74. CURE (Clustering Using REpresentatives) CURE was proposed by Guha, Rastogi & Shim, 1998 It stops the creation of a cluster hierarchy if a level consists of k clusters Each cluster has c representatives (instead of one)  Choose c well scattered points in the cluster  Shrink them towards the mean of the cluster by a fraction of   The representatives capture the physical shape and geometry of the cluster  It can treat arbitrary shaped clusters and avoid single-link effect. Merge the closest two clusters  Distance of two clusters: the distance between the two closest representatives 74 Clustering Analysis
75. 75. CURE Algorithm75 Clustering Analysis
76. 76. CURE Algorithm76 Clustering Analysis
77. 77. CURE Algorithm (Another form)77 Clustering Analysis
78. 78. CURE Algorithm (for Large Databases)78 Clustering Analysis
79. 79. Data Partitioning and Clustering y s = 50 p=2 s/p = 25 x s/pq = 5y y y y x x x x 79 Clustering Analysis
80. 80. Cure: Shrinking Representative Points Shrink the multiple representative points towards the gravity center by a fraction of . Multiple representatives capture the shape of the clustery y x x 80 Clustering Analysis
81. 81. Clustering Categorical Data: ROCK ROCK: RObust Clustering using linKs, by S. Guha, R. Rastogi, K. Shim (ICDE’99).  Use links to measure similarity/proximity  Not distance-based with categorical attributes  Computational complexity: O(n2  nmmma  n2 log n) Basic ideas (Jaccard coefficient): Similarity function and neighbors:  T1  T2 Let T1 = {1,2,3}, T2={3,4,5} Sim(T1 , T2 )  T1  T2 {3} 1Sim(T1, T 2)    0.2 {1,2,3,4,5} 5 81 Clustering Analysis
82. 82. ROCK: An Example Links: The number of common neighbors for the two points. Using Jaccard Use Distances to determine neighbors  (pt1,pt4) = 0, (pt1,pt2) = 0, (pt1,pt3) = 0  (pt2,pt3) = 0.6, (pt2,pt4) = 0.2  (pt3,pt4) = 0.2 Use 0.2 as threshold for neighbors  Pt2 and Pt3 have 3 common neighbors  Pt3 and Pt4 have 3 common neighbors  Pt2 and Pt4 have 3 common neighbors Resulting clusters (1), (2,3,4) which makes more sense 82 Clustering Analysis
83. 83. ROCK: Property & Algorithm Links: The number of common neighbours for the two points. {1,2,3}, {1,2,4}, {1,2,5}, {1,3,4}, {1,3,5} {1,4,5}, {2,3,4}, {2,3,5}, {2,4,5}, {3,4,5} 3 {1,2,3} {1,2,4} Algorithm  Draw random sample  Cluster with links (maybe agglomerative hierarchical)  Label data in disk83 Clustering Analysis
84. 84. CHAMELEON CHAMELEON: hierarchical clustering using dynamic modeling, by G. Karypis, E.H. Han and V. Kumar’99 Measures the similarity based on a dynamic model  Two clusters are merged only if the interconnectivity and closeness (proximity) between two clusters are high relative to the internal interconnectivity of the clusters and closeness of items within the clusters A two phase algorithm 1. Use a graph partitioning algorithm: cluster objects into a large number of relatively small sub-clusters 2. Use an agglomerative hierarchical clustering algorithm: find the genuine clusters by repeatedly combining these sub-clusters 84 Clustering Analysis
85. 85. Graph-based clustering Sparsification techniques keep the connections to the most similar (nearest) neighbors of a point while breaking the connections to less similar points. The nearest neighbors of a point tend to belong to the same class as the point itself. This reduces the impact of noise and outliers and sharpens the distinction between clusters. 85 Clustering Analysis
86. 86. Overall Framework of CHAMELEON Construct Sparse Graph Partition the GraphData Set Merge Partition Final Clusters 86 Clustering Analysis
87. 87. Density-Based Clustering Methods Clustering based on density (local cluster criterion), such as density-connected points Major features:  Discover clusters of arbitrary shape  Handle noise  One scan  Need density parameters as termination condition Several interesting studies:  DBSCAN: Ester, et al. (KDD’96)  OPTICS: Ankerst, et al (SIGMOD’99).  DENCLUE: Hinneburg & D. Keim (KDD’98)  CLIQUE: Agrawal, et al. (SIGMOD’98) 87 Clustering Analysis
88. 88. Density-Based Clustering (Background) Two parameters:  Eps: Maximum radius of the neighbourhood  MinPts: Minimum number of points in an Eps-neighbourhood of that point NEps(p): {q belongs to D | dist(p,q) <= Eps} Directly density-reachable: A point p is directly density -reachable from a point q wrt. Eps, MinPts if 1) p belongs to NEps(q) 2) core point condition: p MinPts = 5 |NEps (q)| >= MinPts q Eps = 1 cm 88 Clustering Analysis
89. 89. Density-Based Clustering (Background) The neighborhood with in a radius Eps of a given object is call the  neighborhood of the object. If the neighborhood of an object contains at least a minimum number, MinPts, of objects, then the object is called a core object. Given a set of objects, D, we say that an object p is directly density- reachable from object q if p is within the neighborhood of q, and q is a core object. An object p is density-reachable from object q with respect to  and MinPts in a set of objects D, if there is a chain of object p1, …, pn, where p1=q and pn=p such that p(i+1) is directly density-reachable from pi with respect to  and MinPts, for 1≤i≤n, from pi  D. An object p is density-connected to object q with respect to 89 Clustering Analysis
90. 90. Density-Based Clustering p p1 q Density-reachable:  A point p is density-reachable from a point q wrt. Eps, MinPts if there is a chain of points p1, …, pn, p1 = q, pn = p such that pi+1 is directly density- reachable from pi Density-connected  A point p is density-connected to a point q wrt. Eps, MinPts if there is a point o such that both, p and q are density-reachable from o wrt. Eps and MinPts. p q o 90 Clustering Analysis
91. 91. DBSCAN: Density Based Spatial Clustering of Applications withNoise Relies on a density-based notion of cluster: A cluster is defined as a maximal set of density-connected points Discovers clusters of arbitrary shape in spatial databases with noise Outlier Border Eps = 1cm Core MinPts = 5 91 Clustering Analysis
92. 92. DBSCAN: The Algorithm  Arbitrary select a point p  Retrieve all points density-reachable from p wrt Eps and MinPts.  If p is a core point, a cluster is formed.  If p is a border point, no points are density-reachable from p and DBSCAN visits the next point of the database.  Continue the process until all of the points have been processed.92 Clustering Analysis