Jin, HuidongHuidongJinWong, Man-LeungMan-LeungWongProf. LEUNG Kwong Sak2023-03-272023-03-272005IEEE Transactions on Pattern Analysis and Machine Intelligence, 2005, vol. 27 (11), pp. 1710 - 171901628828http://hdl.handle.net/20.500.11861/7597The scalability problem in data mining involves the development of methods for handling large databases with limited computational resources such as memory and computation time. In this paper, two scalable clustering algorithms, bEMADS and gEMADS, are presented based on the Gaussian mixture model. Both summarize data into subclusters and then generate Gaussian mixtures from their data summaries. Their core algorithm, EMADS, is defined on data summaries and approximates the aggregate behavior of each subcluster of data under the Gaussian mixture model. EMADS is provably convergent. Experimental results substantiate that both algorithms can run several orders of magnitude faster than expectation-maximization with little loss of accuracy © 2005 IEEE.enLarge DatabaseModel-Based ClusteringClustering AlgorithmComputational ResourcesMixture ModelExpectation MaximizationSubclustersGaussian Mixture ModelAccuracy LossAggregation BehaviorScaling AlgorithmLimited Computational ResourcesCore AlgorithmCluster ScalePosterior ProbabilityCovariance MatrixRegular GridModel AlgorithmClustering TechniquesLog-likelihood ValuesOne-Tailed T-testCovariate InformationVector of CovariatesClustering AccuracyModel-Based AlgorithmAcceleration FactorGaussian DensityModel-Based TechniquesConjugate PriorSufficient StatisticsPeer Reviewed Journal Article10.1109/TPAMI.2005.226