Data Stream MiningExcerpts from Wei Fan et al.'sPAKDD Tutorial Notes
Data Stream Mining Excerpts from Wei Fan et al.’s PAKDD Tutorial Notes
UTDTHEUNIVERSITYOFTEXASATDALLASTutorial: Data Stream MiningChallenges and TechniquesPAKDD201124-27May2011,Shenzhen,ChindLatifurKhan1,WeiFan?,Jiawei Han3,Jing Gao3MohammadMMasud11DepartmentofComputerScience,UniversityofTexasatDallas2IBMT.J.WatsonResearch,USA3Departmentof Computer Science,University of Ilinois atUrbana ChampaignTBM
Tutorial: Data Stream Mining Challenges and Techniques Latifur Khan1 , Wei Fan2 , Jiawei Han3 , Jing Gao3 , Mohammad M Masud1 1Department of Computer Science, University of Texas at Dallas 2 IBM T.J. Watson Research , USA 3Department of Computer Science, University of Illinois at Urbana Champaign
Introduction· Data Stream Classification·Clustering· Novel Class DetectionMay24,2011Khanetal.IBM
Introduction Data Stream Classification Clustering Novel Class Detection I Khan et al. May 24, 2011
IntroductionCharacteristics of Data streams are:ContinuousflowofdatadoExamples:TOlCNNII%A1P1805N:19P90.1151ATIAM.RNetworktrafficwmaI.:SensordataCallcenterrecords
Introduction Characteristics of Data streams are: ◦ Continuous flow of data Network traffic Sensor data Call center records ◦ Examples:
ChallengesInfinite lengthConcept-driftConcept-evolutionFeature Evolution
Infinite length Concept-drift Concept-evolution Feature Evolution Challenges