G. Karypis, E.H. Han, V. Kumar. CHAMELEON: A Hierarchical Clustering Algorithm Using Dynamic Modeling. IEEE Computer, 32(8):68--75, 1999.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....structure and its effective usage by the learners. In order to cluster sessions, after identifying the sessions in a pre processing phase, we used clustering algorithms known for their ability to handle categorical data: ROCK [4] an algorithm that acts on a sample of the dataset, CHAMELEON [5], which is based on graph partitioning, and a new algorithm TURN for discrete distributions that we introduced in [2] All of these algorithms, when used in the past for clustering web sessions, have treated sessions as unordered sets of clicks. The similarity measures used to compare sessions ....

....sessions Sessions in cluster#1 Sessions in cluster#2 Sessions in cluster#3 Figure 3. Session clustering visualization tem web log. Both Jaccard similarity and our DynamicProgramming Based similarity methods were used to provide similarity matrices for the given session set. ROCK[4] CHAMELEON[5] and TURN[2] were then applied on the similarity matrices to each produce clustering result. From the clustering results, we found that ROCK tends to find bigger clusters with lower average similarity. CHAMELEON and TURN can find clusters with high internal cross similarity. The difference between ....

G. Karypis, E.-H. Han, and V. Kumar. Chameleon: A hierarchical clustering algorithm using dynamic modeling. IEEE Computer, 32(8):68--75, August 1999.

....and grid based methods. This section gives a taxonomy analysis and an experimental study of representative methods in each group. In order to examine the clustering ability of clustering algorithms, we performed experimental evaluation upon k means [12] CURE [21] ROCK [8] DBSCAN [2] CHAMELEON [14], WaveCluster [24] and CLIQUE [1] The DBSCAN code came from its authors while CURE and ROCK codes were kindly supplied by the Department of Computer Science and Engineering, University of Minnesota. k means, CHAMELEON, WaveCluster, and CLIQUE programs were locally implemented. We evaluate these ....

....were kindly supplied by the Department of Computer Science and Engineering, University of Minnesota. k means, CHAMELEON, WaveCluster, and CLIQUE programs were locally implemented. We evaluate these algorithms by using two dimensional spatial data sets referenced and used in the CHAMELEON paper [14] and data sets referenced and used in the WaveCluster paper [24] The reason for using two dimensional spatial data is because we can visually evaluate the quality of the clustering result. Often people can intuitively identify clusters on two dimensional spatial data, while this is usually very ....

[Article contains additional citation context not shown here]

Karypis G., Han E.-H. and Kumar V. (1999) Chameleon: A hierarchical clustering algorithm using dynamic modeling. IEEE Computer, 32(8) pp68--75.

.... large genuine clusters, which is the case for partitional approaches such as kmeans [6] failure to handle convex and elongated shapes of clusters as is the case with most hierarchical approaches such as CURE [11] and ROCK [4] and the sensitivity to noise in the data such as in CHAMELEON s case [7] or DBSCAN [2] a density based clustering algorithm. The greatest difficulty in the field of data clustering is the need for input parameters. Many algorithms, especially the hierarchical methods [5, 11] require the initial choice of the number of clusters to find. Even where this is not ....

....called TURN which does not require the input of any parameter and still efficiently discovers clusters of complex shapes in very large data sets. We report the efficiency and demonstrate the effectiveness of TURN on large and complex data sets containing points in 2D space borrowed from [7, 13] for comparison reasons. Section 2 gives an overview of existing, well known clustering algorithms and Section 3 describes our clustering algorithm TURN . Section 4 presents some experimental results for TURN and six established clustering algorithms. Finally, Section 5 concludes and discusses ....

[Article contains additional citation context not shown here]

G. Karypis, E.-H. Han, and V. Kumar. CHAMELEON: A hierarchical clustering algorithm using dynamic modeling. IEEE Computer, 32(8):68--75, August 1999.

....tasks that deal with large numbers of documents. In our experiments, one of the corpora we used is small enough (2745 documents) to allow us to compare CBC with these hierarchical algorithms. Chameleon is a hierarchical algorithm that employs dynamic modeling to improve clustering quality [9]. When merging two clusters, one might consider the sum of the similarities between pairs of elements across the clusters (e.g. average link clustering) A drawback of this approach is that the existence of a single pair of very similar elements might unduly cause the merger of two clusters. An ....

Karypis, G.; Han, E.-H.; and Kumar, V. 1999. Chameleon: A hierarchical clustering algorithm using dynamic modeling. 1EEE Computer: Special Issue on Data Analysis and Mining 32(8):68W5.

....this similarity as the average similarity between all pairs of elements across clusters. The complexity of these algorithms is O(n21ogn) where n is the number of elements to be clustered [6] Chameleon is a hierarchical algorithm that employs dynamic modeling to improve clustering quality [7]. When merging two clusters, one might consider the sum of the similarities between pairs of elements across the clusters (e.g. average link clustering) A drawback of this approach is that the existence of a single pair of very similar elements might unduly cause the merger of two clusters. An ....

Karypis, G.; Han, E.-H.; and Kumar, V. 1999. Chameleon: A hierarchical clustering algorithm using dynamic modeling. IEEE Computer: Special Issue on Data Analysis and Mining 32(8):68 75.

....tasks that deal with large numbers of documents. In our experiments, one of the corpora we used is small enough (2745 documents) to allow us to compare CBC with these hierarchical algo rithms. Chameleon is a hierarchical algorithm that employs dynamic modeling to improve clustering quality [8]. When merging two clusters, one might consider the sum of the similarities between pairs of elements across the clusters (e.g. average link clustering) A drawback of this approach is that the existence of a single pair of very similar elements might unduly cause the merger of two clusters. An ....

Karypis, G.; Han, E.-H.; and Kumar, V. 1999. Chameleon: A hierarchical clustering algorithm using dynamic modeling. IEEE Computer: Special Issue on Data Analysis and Mining 32(8):68 75.

....similarity as the average similarity between all pairs of elements across clusters. The complexity of these algorithms is O(n logn) where n is the number of elements to be clustered [6] Chameleon is a hierarchical algorithm that employs dynamic modeling to improve clustering quality [7]. When merging two clusters, one might consider the sum of the similarities between pairs of elements across the clusters (e.g. average link clustering) A drawback of this approach is that the existence of a single pair of very similar elements might unduly cause the merger of two ....

Karypis, G.; Han, E.-H.; and Kumar, V. 1999. Chameleon: A hierarchical clustering algorithm using dynamic modeling. IEEE Computer: Special Issue on Data Analysis and Mining 32(8):6875.

....is more relevant to investigate clustering algorithms meeting the specific requirement of minimizing the I O operations. Some of the major clustering algorithms proposed in the context of data mining are BIRCH[14] CURE[10] PAM[12] CLARANS[12] DBSCAN[4] BUBBLE[7] MAFIA [8] ITERATE , CHAMELON[11] etc. It is to be noted that the basic principle of clustering hinges on a concept of distance metric or similarity metric. Thus the clustering techniques that are designed mostly for numeric data, exploit the inherent geometric properties based on some priori structure for numerical data. ....

. G. Karyapis, E.H. Han and V. Kumar. CHAMELEON: A hierarchical clustering algorithm using dynamic modeling. In IEEE Computer, Special Issue on Data Analysis and Mining.

....marketing and customer segmentation. Clustering typically groups data into sets in such a way that the intra cluster similarity is maximized while the inter cluster similarity is minimized. Many efficient clustering algorithms, such as ROCK [1] DBSCAN [3] BIRTH [4] C2p [2] CURE [5] CHAMELEON [6], WaveCluster [7] and CLIQUE [8] have been proposed by the database research community. Most previous works in clustering focus on numerical data whose inherent geometric properties can be exploited naturally to define distance functions between data points. However, much of the data existed in ....

G. Karypis, E.-H. Han, V. Kumar: "CHAMELEON: A Hierarchical Clustering Algorithm Using Dynamic Modeling". IEEE Computer, Vol. 32, No. 8, 68-75, 1999.

....single groups have similar characteristics, while dissimilar objects are placed in separate groups. Various existing clustering algorithms have been proposed and designed to fit various formats and constraints of application including k means [1] k medoids [2] BIRCH [3] CURE [4] CHAMELEON [5], DBSCAN [6] AUTOCLUST [7] and AUTOCLUST [8] No single algorithms is suitable for all types of objects, nor are all algorithms appropriate for all problems, however, k medoids algorithms have been shown to be robust to outliers This report is an extension and revision of a paper presented ....

G. Karypis, E.-H. Han, and V. Kumar, "Chameleon: a hierarchical clustering algorithm using dynamic modeling," Computer, vol. 32, pp. 32--68, 1999.

....demonstration of activating the separating operators on two graphs is given in [4] 4. 3 Clustering spatial points We now illustrate the ability of our method to cluster correctly 2D points, in a number of typical cases, some of which have been shown to be problematic for agglomerative methods [6]. More extensive examples are given in Section 6. We show only examples in 2D, although the method works well in higher dimensions too, because two dimensions are easier to visualize and evaluate. We have used 10 mutual neighborhood graphs for modeling the points ( intersection with the ....

....from URL: http: www.cs.cmu.edu quake triangle.html. The reader is encouraged to see the full electronic version of this paper [4] in order to view the figures of this section in larger and clearer format, and in color. Figure 4 shows the results of the algorithm on data sets taken from [6]. These data sets contain clusters of di#erent shapes, sizes and densities and also random noise. A nice property of our algorithm is that random noise gets to stay inside small clusters. After clustering the data, the algorithm treats all the relatively small clusters, whose sizes are below half ....

[Article contains additional citation context not shown here]

G. Karypis, E. Han, and V. Kumar, "CHAMELEON: A Hierarchical Clustering Algorithm Using Dynamic Modeling", IEEE Computer, 32 (1999), 68--75.

....is represented by a tree strategy that separates the objects into small subsets until each subset consists only of suf ciently similar objects. There exist a large number of clustering algorithms in the literature including k means [7] k medoids [8] CACTUS [9] CURE [10] CHAMELEON [11] and DBSCAN [12] No single algorithms is suitable for all types of objects, nor are all algorithms appropriate for all problems, however, the k medoids algorithms have been shown to be robust to outliers [8] compared with centroid based clustering. Assume T objects x 1 ; x 2 ; x T and k ....

G. Karypis, E.-H. Han, and V. Kumar, \Chameleon: a hierarchical clustering algorithm using dynamic modeling," Computer, vol. 32, pp. 32-68, 1999.

....spatial data) the algorithm ROCK developed by the same researchers [GRS99] targets hierarchical agglomerative clustering for categorical attributes. It is surveyed in the section Co Occurrence of Categorical Data. The hierarchical agglomerative algorithm CHAMELEON developed by Karypis et al. [KHK99] utilizes dynamic modeling in cluster aggregation. It uses connectivity graph G corresponding to the k nearest neighbor model sparsification of connectivity matrix: edges of k most similar points are preserved, the rest are pruned. CHAMELEON has two stages. In the first stage small tight clusters ....

Karypis, G., Han, E.-H., and Kumar, V., CHAMELEON: A hierarchical clustering algorithm using dynamic modeling, COMPUTER, 32:68-75, 1999. 47

....well as due to performance restrictions of today s graphics hardware we are forced to the use an efficient level of detail strategy. Consequently, literature describes various interesting data clustering approaches including their efficient and refined implementations [5] 8] 11] 12] 16] [17], 24] Because our main interest lies in visualizing clusters, we focus on the problem of clustering large data sets in coordinate space [7] also referred to as the Euclidian space, in which data objects can be represented as vectors . Unlike data sets in a distance space [7] also referred to ....

....CURE [11] remedies the drawback of single centroid representation by taking advantage of a multi centroid representation of clusters. Hence this algorithm is more robust to outliers and identifies clusters varying in size and having non spherical shapes. A recent approach is called CHAMELEON [17], a hierarchical clustering algorithm that measures inter cluster similarity based on a dynamic model. In addition to other algorithms, CHAMELEON clustering is based not only on vicinity of objects but also considers corresponding connectivity information. This combination results in a robust ....

G. Karypis, Eui-Hong Han, V. Kumar. "CHAMELEON: A Hierarchical Clustering Algorithm Using Dynamic Modeling. " In IEEE Computer, Volume: 32 Issue: 8 , pp. 68-75, Aug. 1999.

....sampling and partitioning to detect clusters. ROCK [11] presents a technique for clustering boolean and categorical data. DBSCAN [6] uses density measure to detect clusters. BIRCH [29] presents a hierarchical clustering based approach that minimizes computational and space requirements. CHAMELEON [14] employs graph partitioning to detect clusters in a hierarchical fashion. WaveCluster [25] uses wavelets to detect clusters in a spatial dataset. DBCLASD [28] uses a distribution based approach for cluster detection in spatial datasets. CLARANS [20] uses randomized search for cluster detection. ....

G. Karypis, E.-H. Han, and V. Kumar. Chameleon: A hierarchical clustering algorithm using dynamic modeling. IEEE Computer, page to appear.

....algorithms according to categorical data. As we have mentioned above we are going to use categorical attributes in order to characterize the documents. An additional important factor is the concepts of Relative Closeness (RC) and Relative Interconnectivity (RI) presented in Chameleon algorithm [16]. Existing clustering algorithms find clusters that fit some static model. Although effective in some cases, these algorithms can break down; that is, cluster the data incorrectly, if the user does not select appropriate static model parameters. The problem is that they use this static model of ....

Karypis G., Eui-Hong (Sam) Han and Vipin Kumar, "Chameleon: Hierarchical Clustering Algorithm Using Dynamic Modeling", Computer Magazine, August 1999.

....version of CLIQUE, and handles both large size and high dimensionality well. The computation time complexity for the revised algorithm is O(c k ) where c is a constant and k is the distinct number of dimensions represented by the cluster subspaces in the dataset. 6. 3 CHAMELEON CHAMELEON [KHK99] combines a graph partitioning algorithm with a hierarchical clustering scheme that dynamically creates clusters. The first step of the algorithm partitions the data using a method based on the k nearest neighbor approach to graph partitioning. In the graph, the density of a region is stored as ....

Karypis, George, Eui-Hong Han, and Vipin Kumar. (1999). CHAMELEON: A Hierarchical Clustering Algorithm Using Dynamic Modeling. IEEE Computer, 32:8, August, 1999. pp. 68-75.

....in different areas including data mining. Surveys on clustering algorithms can be found in [1, 9] Among the more prominent clustering algorithms hierarchical clustering is one. It was first proposed in 1951 [6] and since then there have been numerous modifications including the recent ones [8, 10]. The fact that it has been a prominent algorithm for last half a century shows that it has stood the test of time. Among its various good features it is a non parametric (assumes very little in the way of data characteristics) natural and simple way of grouping objects, and capable of finding ....

....is the way out. Type (1) algorithm of previous paragraph that uses priority queues is a stored matrix method while type (2) algorithm is more suitable as stored data if M is not large and if memory is not enough to store O(N 2 ) similarity matrix. Recent algorithms on hierarchical clustering [8, 10, 12] uses priority queues. The algorithms presented here improves both stored matrix and stored data algorithms by reducing their CPU time significantly, and furthermore it reduces the memory requirement substantially for stored matrix algorithm. 3 Proposed Algorithms In the previous section we ....

[Article contains additional citation context not shown here]

G. Karypis, E-H. Han, and V. Kumar. CHAMELEON: A hierarchical clustering algorithm using dynamic modeling. IEEE Computer, 32:68--75, 1999.

....data mining tools. Clustering is fundamental to nd spatial concentrations. In particular, for spatial analysis, clusters are aggregations resulting from spatial interactions between neighbors, thus cluster identi cation is of great importance for further analysis. Various clustering methods [4, 5, 6, 7, 8, 9, 12, 13, 16, 18, 20, 21] have been proposed for the spatial databases manipulated in GIS. However, most exploratory and robust spatial clustering approaches deal with two dimensional data and do not easily extend to three dimensional data. In many situations, three dimensional data models real world phenomenon better ....

....approaches, naturally in 2D [1, 4, 7, 8, 9, 11] Since the Short Long analysis of edges originated from Delaunay Triangulations, the three dimensional counterpart (Delaunay Tetrahedrization) is a natural candidate for threedimensional modeling of spatial proximity. On the other hand, CHAMELEON [13] successfully derives clusters from k nearest neighbor graphs. However, CHAMELEON [13] uses argumentspeci c neighboring that represents non symmetric adjacency. This occurs when p i is a neighbor of p j , but p j is not a neighbor of p i . This non symmetry is inappropriate for graph based ....

[Article contains additional citation context not shown here]

G. Karypis, E. Han, and V. Kumar. CHAMELEON: A Hierarchical Clustering Algorithm Using Dynamic Modeling. IEEE Computer: Special Issue on Data Analysis and Mining, 32(8):68-75, 1999.

....of several input parameters. Distance between clusters is evaluated by multiple representative points which is well suited for arbitrary shaped clusters. CHAMELEON is a recently proposed hierarchical clustering algorithm which performs well for arbitrary shaped clusters in 2 dimensional spaces [KHK99] 9 2.1.2 Partition based Clustering Partitioning algorithms construct a partition of the objects in the database into clusters such that objects in a cluster are more similar to each other than to the objects in different clusters. k means and k medoid methods determine k cluster ....

G. Karyapis, E.H. Han, and V. Kumar. CHAMELEON: A hierarchical clustering algorithm using dynamic modeling. IEEE Computer, Special Issue on Data Analysis and Mining, 1999.

No context found.

G. Karypis, E.H. Han, and V. Kumar. Chameleon: A hierarchical clustering algorithm using dynamic modeling. IEEE Computer, 32(8):68--75, 1999.

No context found.

G. Karypis, E.H. Han, and V. Kumar. Chameleon: A hierarchical clustering algorithm using dynamic modeling. IEEE Computer, 32(8):68--75, 1999.

No context found.

G. Karypis, E.H. Han, and V. Kumar. Chameleon: A hierarchical clustering algorithm using dynamic modeling. IEEE Computer, 32(8):68--75, 1999.

No context found.

G. Karypis, E.H. Han, and V. Kumar. Chameleon: A hierarchical clustering algorithm using dynamic modeling. IEEE Computer, 32(8):68--75, 1999.

No context found.

G. Karypis, E.H. Han, and V. Kumar. Chameleon: A hierarchical clustering algorithm using dynamic modeling. IEEE Computer, 32(8):68--75, 1999.

No context found.

G. Karypis, E.H. Han, and V. Kumar. Chameleon: A hierarchical clustering algorithm using dynamic modeling. IEEE Computer, 32(8):68--75, 1999.

No context found.

G. Karypis, E.H. Han, and V. Kumar. Chameleon: A hierarchical clustering algorithm using dynamic modeling. IEEE Computer, 32(8):68--75, 1999.

....Keywords Partitioning, Maximum degree, Placement, Congestion 1. INTRODUCTION Hypergraph partitioning is an important problem with extensive applications to many areas, including VLSI design [5] ecient storage of large databases on disks [23] information retrieval [28] and data mining [9, 15]. The problem is to partition the vertices of a hypergraph into k equal size subdomains, such that the number of the hyperedges connecting vertices in di erent subdomains (called the cut) is minimized. This work was supported by NSF ACI 0133464, CCR9972519, EIA 9986042, ACI 9982274, and by Army ....

G. Karypis, E. Han, and V. Kumar. Chameleon: A hierarchical clustering algorithm using dynamic modeling. IEEE Computer, 32(8):68-75, 1999.

.... MN 55455 karypis cs.umn.edu Introduction Hypergraph partitioning is an important problem with extensive application to many areas, including VLSI design [Alpert and Kahng, 1995] efficient storage of large databases on disks [Shekhar and Liu, 1996] and data mining [Mobasher et al. 1996, Karypis et al. 1999b] The problem is to partition the vertices of a hypergraph into k equal size parts, such that the number of hyperedges connecting vertices in different parts is minimized. During the course of VLSI circuit design and synthesis, it is important to be able to divide the system specification into ....

.... developed for partitioning large graphs derived from scientific computations [Hendrickson and Leland, 1994, Karypis and Kumar, 1998b, Walshaw et al. 1997] but their advantages were quickly recognized by the VLSI CAD community, and a number of different multilevel algorithms have been developed [Karypis et al. 1999a, Alpert et al. 1997, Karypis and Kumar, 2000] In this chapter we try to provide an overview of the multilevel paradigm and describe the various algorithms that it uses and why it works. Even though our presentation will be generic, from time to time we will use our experience in developing the ....

[Article contains additional citation context not shown here]

Karypis, G., Han, E., and Kumar, V. (1999b). Chameleon: A hierarchical clustering algorithm using dynamic modeling. IEEE Computer, 32(8):68--75.

.... many times when clusters have subclusters, and the hierarchical structure is indeed a natural constrain on the underlying application domain (e.g. biological taxonomies, phylogenetic trees, etc) 14] Hierarchical clustering solutions have been primarily obtained using agglomerative algorithms [35, 23, 15, 16, 21], in which objects are initially assigned to their own cluster and then pairs of clusters are repeatedly merged until the whole tree is formed. However, partitional algorithms [27, 20, 29, 6, 42, 19, 37, 5, 13] can also be used to obtain hierarchical clustering solutions via a sequence of repeated ....

....algorithms is the method used to determine the pair of clusters to be merged at each step. In most agglomerative algorithms, this is accomplished by selecting the most similar pair of clusters, and numerous approaches have been developed for computing the similarity between two clusters[35, 23, 20, 15, 16, 21]. In our study we used the single link, complete link, and UPGMA schemes as well as the various partitional criterion functions described in Section 3.1. The single link [35] scheme measures the similarity of two clusters by the maximum similarity between the documents from each cluster. That is, ....

G. Karypis, E.H. Han, and V. Kumar. Chameleon: A hierarchical clustering algorithm using dynamic modeling. IEEE Computer, 32(8):68--75, 1999.

.... are many times when clusters have subclusters, and the hierarchical structure are indeed a natural constrain on the underlying application domain (e.g. biological taxonomy, phylogenetic trees) 9] Hierarchical clustering solutions have been primarily obtained using agglomerative algorithms [27, 19, 10, 11, 18], in which objects are initially assigned to its own cluster and then pairs of clusters are repeatedly merged until the whole tree is formed. However, partitional algorithms [22, 16, 24, 5, 33, 13, 29, 4, 8] can also be used to obtain hierarchical clustering solutions via a sequence of repeated ....

....algorithms is the method used to determine the pairs of clusters to be merged at each step. In most agglomerative algorithms, this is accomplished by selecting the most similar pair of clusters, and numerous approaches have been developed for computing the similarity between two clusters[27, 19, 16, 10, 11, 18]. In our study we used the single link, complete link, and UPGMA schemes, as well as, the various partitional criterion functions described in Section 3.1. The single link [27] scheme measures the similarity of two clusters by the maximum similarity between the documents from each cluster. That ....

G. Karypis, E.H. Han, and V. Kumar. Chameleon: A hierarchical clustering algorithm using dynamic modeling. IEEE Computer, 32(8):68--75, 1999.

....when the clusters are of different tightness, and (ii) the degree to which they can lead to reasonably balanced clusters. I Introduction The topic of clustering has been extensively studied in many scientific disciplines and over the years a variety of different algorithms have been developed [31, 22, 6, 27, 20, 35, 2, 48, 13, 43, 14, 15, 24]. Two recent surveys on This work was supported by NSF CCR 9972519, EIA 9986042, ACI 9982274, by Army Research Office contract DA DAAG55 98 1 0441, by the DOE ASCI program, and by Army High Performance Computing Research Center contract number DAAH04 95 C 0008. Related papers are available via ....

....repeatedly merging pairs of clusters until a certain stopping criterion is met. A number of different methods have been proposed for determining the next pair of clusters to be merged, such as group average (UPGMA) 22] single link [38] complete link [28] CURE [14] ROCK [15] and CHAMELEON [24]. Hierarchical algorithms produce a clustering that forms a dendrogram, with a single all inclusive cluster at the top and single point clusters at the leaves. On the other hand, partitional algorithms, such as K means [33, 22] K medoids [22, 27, 35] Autoclass [8, 6] graph partitioning based ....

[Article contains additional citation context not shown here]

G. Karypis, E.H. Hun, and V. Kumar. Chameleon: A hierarchical clustering algorithm using dynamic modeling. IEEE Computer, 32(8):685, 1999. 27

....when the clusters are of different tightness, and (ii) the degree to which they can lead to reasonably balanced clusters. 1 Introduction The topic of clustering has been extensively studied in many scientific disciplines and over the years a variety of different algorithms have been developed [31, 22, 6, 27, 20, 35, 2, 48, 13, 43, 14, 15, 24]. Two recent surveys on # This work was supported by NSF CCR 9972519, EIA 9986042, ACI 9982274, by Army Research Office contract DA DAAG55 98 1 0441, by the DOE ASCI program, and by Army High Performance Computing Research Center contract number DAAH04 95 C 0008. Related papers are available via ....

....repeatedly merging pairs of clusters until a certain stopping criterion is met. A number of different methods have been proposed for determining the next pair of clusters to be merged, such as group average (UPGMA) 22] single link [38] complete link [28] CURE [14] ROCK [15] and CHAMELEON [24]. Hierarchical algorithms produce a clustering that forms a dendrogram, with a single all inclusive cluster at the top and single point clusters at the leaves. On the other hand, partitional algorithms, such as K means [33, 22] K medoids [22, 27, 35] Autoclass [8, 6] graph partitioning based ....

[Article contains additional citation context not shown here]

G. Karypis, E.H. Han, and V. Kumar. Chameleon: A hierarchical clustering algorithm using dynamic modeling. IEEE Computer, 32(8):68--75, 1999. 27

....and G is a nite clustering space, it is impossible to have the DISP value decreasing forever. Thus, the algorithm terminates. 6 Scaling the Algorithm by Micro cluster Sharing For clustering large, disk resident databases, many studies have adopted a microclustering methodology (e.g. [ZRL96,WYM97,BFR98,KHK99]) which compresses data objects into micro clusters in a pre clustering phase so that the subsequent clustering activities can be accomplished at the micro cluster level. To ensure that not much quality is lost, a maximum radius on a micro cluster is imposed. By micro clustering, in our ....

G. Karypis, E.-H. Han, and V. Kumar. CHAMELEON: A hierarchical clustering algorithm using dynamic modeling. COMPUTER, 32:68-75, 1999.

No context found.

G. Karypis, E.H. Han, V. Kumar. CHAMELEON: A Hierarchical Clustering Algorithm Using Dynamic Modeling. IEEE Computer, 32(8):68--75, 1999.

No context found.

G. Karypis, E.-H. Han, and V. Kumar. 1999. Chameleon: A Hierarchical Clustering Algorithm using Dynamic Modeling. IEEE Computer: Special Issue on Data Analysis and Mining, 32(8):68--75.

No context found.

KARYPIS G. etc. CHAMELEON: A Hierarchical Clustering Algorithm Using Dynamic Modeling. In: Computer, 32: 68-75, 1999.

No context found.

G. Karypis, E-H Han, and V. Kumar. CHAMELEON: A hierarchical clustering algorithm using dynamic modeling. IEEE Computer, 32(8):68--75, 1999.

No context found.

G. Karypis, E.-H. Han, and V. Kumar. Chameleon: A Hierarchical Clustering Algorithm Using Dynamic Modeling. IEEE Computer, 32(8):68--75, 1999.

No context found.

G. Karypis, E.-H. Han, and V. Kumar. Chameleon: A Hierarchical Clustering Algorithm Using Dynamic Modeling. IEEE Computer, 32(8):68--75, 1999.

No context found.

G. Karypis, E. H. Han, and V. Kumar. Chameleon: A hierarchical clustering algorithm using dynamic modeling. COMPUTER, 32, 1999.

No context found.

George Karypis, Eui-Hong Han, and Vipin Kumar. Chameleon: A hierarchical clustering algorithm using dynamic modeling. IEEE Computer, 32(8):68--75, August 1999.

No context found.

G. Karypis, E.H. Han, and V. Kumar, "CHAMELEON:a hierarchical clustering algorithm using dynamic modeling," IEEE Computer, 1999.

No context found.

Karypis, G., Han, E.H. and Kumar, V. 1999. CHAMELEON: a hierarchical clustering algorithm using dynamic modeling. COMPUTER, 32:68--75.

No context found.

G. Karypis, E.-H. Han and V. Kumar, "Chameleon: A hierarchical clustering algorithm using dynamic modeling", IEEE Computer: Special Issue on Data Analysis and Mining, Vol. 32, No. 8, 1999.

No context found.

G. Karypis, E.-H. Han, and V. Kumar, "CHAMELEON: a hierarchical clustering algorithm using dynamic modeling," Computer, vol. 32, pp. 32--68, 1999.

No context found.

G. Karyapis, E.H. Han and V. Kumar. CHAMELEON: A Hierarchical Clustering Algorithm Using Dynamic Modeling, In IEEE Computer, Special Issue on Data Analysis and Mining.

No context found.

G. Karypis, E.-H. Han, and V. Kumar. Chameleon: A Hierarchical Clustering Algorithm Using Dynamic Modeling. IEEE Computer, 32(8):68--75, 1999.

No context found.

G. Karypis, E. Han, and V. Kumar. CHAMELEON: A Hierarchical Clustering Algorithm Using Dynamic Modeling. IEEE Computer: Special Issue on Data Analysis and Mining, 32(8):68-75, 1999.

No context found.

G. Karypis, E. Han, and V. Kumar. Chameleon: A hierarchical clustering algorithm using dynamic modeling. In IEEE Computer, pages 68--75, 1999.

First 50 documents  Next 50

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC