Transcription
In-Database Analytics with RMichele Chambers – Advanced Analytics Product Management, DirectorBrian Hess – Advanced Analytics, Director & Principal Mathematician 2010 Netezza, Inc. All rights reserved
Agenda What are in-database analytics? How does in-database analytics processing help you? Can in-database analytics be used for data mining as well as scoring? How can you take advantage of a massively parallel architecture to speed upembarrassingly parallel algorithms as well as heroic computations?Page 2Use R! 2010 - Netezza - In-database Analytics with R
Advanced Analytics – the Traditional astingSQLETLSQLR, S FraudDetectionETLC/C , Java, Python,Fortran, SQLPage 3Use R! 2010 - Netezza - In-database Analytics with R
What are in-database analytics?Embedding of analytics inside the databaseso that the computation processing occurs asclose to the data as possiblePage 4Use R! 2010 - Netezza - In-database Analytics with R
What’s the Big Deal with In-Database Analytics?FPGACPUR ClientAnalyticsBI nClientLoaderAnalyticsDiskEnclosuresS-Blades NetworkFabricApplicationsNetezza AppliancePage 5Use R! 2010 - Netezza - In-database Analytics with R
Moving Compute Next to Data As Data Streams ByTask Parallelism Data ParallelismModel Simulation /Experimentation Concurrent simulationon different data 100’s different modelsrunning against 1M‘srows Scoring / Predicting Concurrent calculationof a prediction or scoreon large quantities ofdataData Mining withTask/Data Parallelism Series of iterations thatcan be parallelizedDefine problemPrepare dataMonte Carlo SimulationControl ProgramSimulation#1on Core 1Simulation#Xon Core XScoring ModelControl ProgramScorecalculationon Core 1Scorecalculationon Core XExplore dataBuild modelValidate modelDeploy and use modelPage 6Use R! 2010 - Netezza - In-database Analytics with R
InIn-Database AnalyticsData snzMatrixScientificAnalyticsOpen Source AnalyticsCustomer/PartnerAnalyticsCustomSoftware Development KitParallel Analytic EnginesnzEnginefor HadoopRAnalyticsnzEngineforRnzAdaptorsforC, C , Java,Python, FortrannzPlug-inforEclipsenzPackageforR GUIStreaming AcceleratorNetezza AMPP PlatformPage 7Use R! 2010 - NetezzaCompany- In-databaseConfidentialAnalytics with R
What does In-Database Analytics yFasterturnarounds 010101010101010Ability toexperimentAbility to ge 8Use R! 2010 - Netezza - In-database Analytics with R
So, What Should I Look For in a Database?In-database analytics checklist1.Data streaming2.Flexible, easy-to-use in-database mechanisms3.Easy, fast, extensible development environment4.Wide availability of tools including open source tools5.Industry accepted standards/tools6.Easy to manage and maintainPage 9Use R! 2010 - Netezza - In-database Analytics with R
Thank youMichele Chambers mchambers@netezza.com 508.382.8264Brian Hess bhess@netezza.com 508.382.8471Page 10Use R! 2010 - Netezza - In-database Analytics with R
What's the Big Deal with In-Database Analytics? R Client BI Client FPGA CPU FPGA Analytics CPU . In-database analytics checklist 1. Data streaming 2. Flexible, easy-to-use in-database mechanisms . Microsoft PowerPoint - In-database Analytics with R v3.pptx
