Filters








711,195 Hits in 6.3 sec

An Experimental Comparison of Several Clustering and Initialization Methods [article]

Marina Meila, David Heckerman
2015 arXiv   pre-print
In the first part of the paper, we perform an experimental comparison between three batch clustering algorithms: the Expectation-Maximization (EM) algorithm, a winner take all version of the EM algorithm  ...  The initializations that we consider are (1) parameters sampled from an uninformative prior, (2) random perturbations of the marginal distribution of the data, and (3) the output of hierarchical agglomerative  ...  Acknowledgments We thank Max Chickering, Chris Meek, and Bo Thiesson for their assistance with the implementation of the algorithms and for many useful and interesting discussions.  ... 
arXiv:1301.7401v2 fatcat:7z47t6t32beuxmkuh6aroqxtb4

Integration Analysis of Diverse Genomic Data Using Multi-clustering Results [chapter]

Hye-Sung Yoon, Sang-Ho Lee, Sung-Bum Cho, Ju Han Kim
2006 Lecture Notes in Computer Science  
In recent years, various methods for ensemble selection and clustering result combinations have been designed to optimize clustering results.  ...  Therefore, a new paradigm is required that combines the genome-wide experimental results of multi-source datasets.  ...  A comparison of the clustering algorithms H50.Low Cluster k -means Hierarchical SOMs Our method Actual value Table 3 . 3 A comparison of the microarray and proteomics datasetsDiversity-based  ... 
doi:10.1007/11946465_4 fatcat:kvhafhaeyfc3lmx6hmraoapiua

A novel framework of the fuzzy c-means distances problem based weighted distance [article]

Andy Arief Setyawan, Ahmad Ilham
2019 arXiv   pre-print
The experimental result using the UCI data set show the proposed method is superior to the original method and other clustering methods.  ...  Clustering is one of the major roles in data mining that is widely application in pattern recognition and image segmentation.  ...  Each method is run 10 times in each dataset, then the average value of the iteration results and the computational times in seconds is taken to obtain an objective comparison of the scores.  ... 
arXiv:1907.13513v1 fatcat:mwiyr57xezcy5b7tnn7mhui7du

The search for experimental design with tens of variables: Preliminary results

Yaileen M. Mendez-Vazquez, Kasandra L. Ramirez-Rojas, Mauricio Cabrera-Rios
2013 2013 Winter Simulations Conference (WSC)  
Several strategies are contrasted: (i) generate designs with random numbers, (ii) use designs already in the literature, and (iii) generate designs under a clustering strategy.  ...  Statistical experimental designs, however, are still somewhat focused on the variation of less than about a dozen variables.  ...  The method starts with an initial experimental design, which for twenty variables has 232 experimental runs using the modified version of the clustering design method.  ... 
doi:10.1109/wsc.2013.6721637 dblp:conf/wsc/Mendez-VazquezRC13 fatcat:d7loelmezremjptfxdlbpoemfy

Max stable set problem to found the initial centroids in clustering problem

Awatif Karim, Chakir Loqman, Youssef Hami, Jaouad Boumhidi
2022 Indonesian Journal of Electrical Engineering and Computer Science  
The latter is sensitive to the random selection of the k cluster centroids in the initialization phase.  ...  methods.  ...  Thus, we propose a method for the automatic detection of initial cluster centroids, which are the input parameters in several partitioning clustering methods.  ... 
doi:10.11591/ijeecs.v25.i1.pp569-579 fatcat:bjttrtgacjhtbacnvfaf3nqyuq

Fuzzy Clustering Algorithm Efficient Implementation Using Centre of Centres

Chundru Ramesh, Komaragiri Rao, Gunamani Jena
2018 International Journal of Intelligent Engineering and Systems  
The experimental research was performed on the publicly available database (i.e. yeast dataset) to validate its clustering performance in terms of accuracy, specificity, sensitivity and execution time.  ...  Clustering is a procedure of finding similar data items (patterns, documents etc.) and then group the similar data together.  ...  The graphical comparison of existing and proposed method using FCM clustering algorithm is represented in the Fig. 3 .  ... 
doi:10.22266/ijies2018.1031.01 fatcat:krmdw6a5o5bybl35x5axblsdcy

Cluster Size Statistic and Cluster Mass Statistic: Two Novel Methods for Identifying Changes in Functional Connectivity Between Groups or Conditions

Alex Ing, Christian Schwarzbauer, Daniele Marinazzo
2014 PLoS ONE  
This pattern of connectivity is known as the functional connectome. Connectivity is often compared between experimental groups and conditions.  ...  Functional connectivity has become an increasingly important area of research in recent years.  ...  Both methods involve an initial comparison between connectivity matrices, giving a matrix of test statistics.  ... 
doi:10.1371/journal.pone.0098697 pmid:24906136 pmcid:PMC4048154 fatcat:qndl5lmfgjharext3q3palm4wm

GAC-GEO: a generic agglomerative clustering framework for geo-referenced datasets

Rachsuda Jiamthapthaksin, Christoph F. Eick, Seungchan Lee
2010 Knowledge and Information Systems  
We evaluate the proposed framework on an artificial dataset and two real world applications involving region discovery.  ...  Major challenges of clustering geo-referenced data include identifying arbitrarily shaped clusters, properly utilizing spatial information, coping with diverse extrinsic characteristics of clusters and  ...  Acknowledgements This research was supported in part by a grant from the Environmental Institute of Houston (EIH).  ... 
doi:10.1007/s10115-010-0355-3 fatcat:agwlvb7wffeczk5fd2lojknedq

A Hierarchical Document Clustering Approach with Frequent Itemsets

Cheng-Jhe Lee, Chiun-Chieh Hsu, Da-Ren Chen
2017 International Journal of Engineering and Technology  
The experimental results reveal that our method is more effective than the well-known document clustering algorithms.  ...  Many conventional document clustering methods perform inefficiently for large document of collected information and require special handling for high dimensionality and high volume.  ...  The accuracies of our method (OCFI) are shown in Table II and Table III , and the comparison of our results with those of several popular clustering algorithms is shown in Table IV .  ... 
doi:10.7763/ijet.2017.v9.965 fatcat:ft2h5bjxj5es3dha3ovydcvoye

Learning Asymmetric Co-Relevance

Fiana Raiber, Oren Kurland, Filip Radlinski, Milad Shokouhi
2015 Proceedings of the 2015 International Conference on Theory of Information Retrieval - ICTIR '15  
Empirical evaluation demonstrates the merits of using the co-relevance estimate in various applications, including cluster-based and graph-based document retrieval.  ...  The model uses different types of similarities with the assumed relevant document and the query, as well as document-quality measures.  ...  Acknowledgments We thank the reviewers for their comments and Kripabandhu Ghosh for initial discussions.  ... 
doi:10.1145/2808194.2809454 dblp:conf/ictir/RaiberKRS15 fatcat:ywfesjip2fedzmdzhmwcw6j254

Unsupervised and Semi-Supervised Clustering for Large Image Database Indexing and Retrieval

Hien Phuong Lai, Muriel Visani, Alain Boucher, Jean-Marc Ogier
2012 2012 IEEE RIVF International Conference on Computing & Communication Technologies, Research, Innovation, and Vision for the Future  
In this article, we present both formal and experimental comparisons of different unsupervised clustering methods for structuring large image databases.  ...  Moreover, a summary of semi-supervised clustering methods is presented and an interactive semi-supervised clustering model using the HMRF-kmeans is experimented on the Wang image database in order to analyse  ...  Experimental comparison In this section, we present an experimental comparison of the partitioning method global k-means [3] with three hierarchical methods (AHC [4] , SR-tree [5] and BIRCH [6] )  ... 
doi:10.1109/rivf.2012.6169869 dblp:conf/rivf/LaiVBO12 fatcat:ewp2bboq4bgpdjfp47ovxrzrr4

A new DSM clustering algorithm for linkage groups identification

Amin Nikanjam, Hadi Sharifi, B. Hoda Helmi, Adel Rahmani
2010 Proceedings of the 12th annual conference on Genetic and evolutionary computation - GECCO '10  
Linkage learning has been considered as an influential factor in success of genetic and evolutionary algorithms for solving difficult optimization problems.  ...  The proposed technique is tested on several benchmark problems and it is shown that it can accurately identify all the linkage groups by O(n 1.7 ) fitness evaluations, where n is problem size.  ...  The proposed clustering strategy is based on natural groups of variables that were seen in the DSM and comprised of several phases to generate and revise these initial groups.  ... 
doi:10.1145/1830483.1830552 dblp:conf/gecco/NikanjamSHR10 fatcat:7xn42r6xw5d75lifk47vfugs7e

Improved$hboxK$-Means Clustering Algorithm for Exploring Local Protein Sequence Motifs Representing Common Structural Property

W. Zhong, G. Altun, R. Harrison, P.C. Tai, Y. Pan
2005 IEEE Transactions on Nanobioscience  
The new initialization method tries to choose suitable initial points, which are well separated and have the potential to form high-quality clusters.  ...  Careful comparison of sequence motifs obtained by the improved and traditional algorithms also suggests that the improved K-means clustering algorithm may discover some relatively weak and subtle sequence  ...  Our experimental results show an average of 40 clusters out of 800 clusters is empty after the first iteration of the traditional K-means algorithm with random selection of initial points.  ... 
doi:10.1109/tnb.2005.853667 pmid:16220690 fatcat:iu3y7juljrbbxg5z4g7gklzsuu

Comparison of FCM and FISODATA

B. Fergani, Mohamed-Khireddine Kholladi, M. Bahri
2012 International Journal of Computer Applications  
In fuzzy clustering, the fuzzy c-means (FCM) clustering algorithm is the best known and used method.  ...  An interesting extension of FCM is the fuzzy ISODATA (FISODATA) algorithm; it updates cluster number during the algorithm. That's why we can have more or less clusters than the initialization step.  ...  Fig 3 : 3 Execution times of both initializations. Fig 4: MSE of both initializations Fig 5 : 2 Fig 6 :Fig 7 : 5267 FISODATA's clustering with m=1.Comparison of FCM's and FISODATA's clustering.  ... 
doi:10.5120/8913-2960 fatcat:sxsuepi45jfqdcqf3zgbzjo3va

Improved KNN Algorithm Based on Preprocessing of Center in Smart Cities

Haiyan Wang, Peidi Xu, Jinghua Zhao, Zhihan Lv
2021 Complexity  
The algorithm can select the center of the spherical region appropriately and then construct an initial classifier for the training set to improve the accuracy and time of classification.  ...  Then, based on it and spherical region division, an improved KNNPK+ is proposed.  ...  Table 1 shows the comparison experimental results of classification accuracy, and Table 2 shows the comparison experimental results of classification time.  ... 
doi:10.1155/2021/5524388 fatcat:tsdfh5va3jhsrg7tl7fhesxkvm
« Previous Showing results 1 — 15 out of 711,195 results