technology update

43
Copyright © 2011, SAS Institute Inc. All rights reserved. Technology Update Deepak Ramanathan Asia Pacific - Technology

Upload: sas-singapore-institute-pte-ltd

Post on 22-Dec-2014

1.991 views

Category:

Technology


3 download

DESCRIPTION

SAS Business Analytics 2011 - Technology Update

TRANSCRIPT

  • 1. Technology UpdateDeepak RamanathanAsia Pacific - TechnologyCopyright 2011, SAS Institute Inc. All rights reserved.

2. 2Copyright 2011, SAS Institute Inc. All rights reserved. 3. Technology Update? 3 Copyright 2011, SAS Institute Inc. All rights reserved. 4. 4Copyright 2011, SAS Institute Inc. All rights reserved. 5. 5Copyright 2011, SAS Institute Inc. All rights reserved. 6. Data Integration Studio 4.36Copyright 2011, SAS Institute Inc. All rights reserved. 7. Whats New in DI Studio 4.3? New and Enhanced Access Engines Greenplum, Sybase IQ, Aster Data7Copyright 2011, SAS Institute Inc. All rights reserved. 8. Whats New in DI Studio 4.3? New Transformations SQL Union Easy import of UDFs into expressions Optimized Oracle Bulk Loader Additional DBMS Expression Builder functions8Copyright 2011, SAS Institute Inc. All rights reserved. 9. Metadata Search9Copyright 2011, SAS Institute Inc. All rights reserved. 10. Metadata Search10Copyright 2011, SAS Institute Inc. All rights reserved. 11. SAS Code Importer Support optional Macro expansion New Analyze only feature Detailed summary information Collision detection with warnings Ex: Library already exists but different path Better logging and error checking Work tables can be registered as physical tables Many more fixes and user requested enhancements11Copyright 2011, SAS Institute Inc. All rights reserved. 12. SAS Code Importer Enhancements 12 Copyright 2011, SAS Institute Inc. All rights reserved. 13. 13Copyright 2011, SAS Institute Inc. All rights reserved. 14. And much more! 14 Copyright 2011, SAS Institute Inc. All rights reserved. 15. 15Copyright 2011, SAS Institute Inc. All rights reserved. 16. Whats New in Reporting Enterprise Guide 5.1 Add in for Microsoft Office 5.1 Roambi for SAS SAS VAE 16 Copyright 2011, SAS Institute Inc. All rights reserved. 17. Enterprise Guide 5.1 17 Copyright 2011, SAS Institute Inc. All rights reserved. 18. EG 5.1 18 Copyright 2011, SAS Institute Inc. All rights reserved. 19. Add-In for MS Office 5.1 - OutlookIntegration19Copyright 2011, SAS Institute Inc. All rights reserved. 20. Add-In for MS Office 5.1 - Task Gallery20Copyright 2011, SAS Institute Inc. All rights reserved. 21. 21Copyright 2010, SAS Institute Inc. All rights reserved. 22. 22Copyright 2011, SAS Institute Inc. All rights reserved. 23. BI Dashboard 4.323Copyright 2011, SAS Institute Inc. All rights reserved. 24. 24Copyright 2011, SAS Institute Inc. All rights reserved. 25. SAS Analytics Whats New Enterprise Miner Model Manager SAS/STAT SAS/ETS SAS/OR 25 Copyright 2011, SAS Institute Inc. All rights reserved. 26. SAS Enterprise Miner 7.1Data MiningRelease Objectives Application tools In-database processingKey Features Survival data mining for time dependent outcomes High performance rate making (claims, severity, and pure premium) Support for logistic regression models in LAR/LASSO node Optimal binning with constraints for credit score carding In-database expansions for statistical transformations and modeling Time series data mining tools (Experimental)26These materials are proprietary to SAS and constitute SAS confidential information pursuant to the terms of the confidentiality or nondisclosure agreement in place between you and SASCopyright 2011, SAS Institute Inc. All rights reserved. 27. SAS Model Manager 3.1 Data Mining Release Objectives Deeper support for existing models Broader support for additional model types Key Features Workflow to replace lifecycle templates and enable automated and collaborative model import, comparison, validation, scoring, monitoring and retraining Model retraining in MM New report types Model management for additional types of models including R models27These materials are proprietary to SAS and constitute SAS confidential information pursuant to the terms of the confidentiality or nondisclosure agreement in place between you and SASCopyright 2011, SAS Institute Inc. All rights reserved. 28. SAS/STAT 9.3StatisticsKey Features New FMM procedure (experimental) for finite mixture modeling Shared frailty models in PHREG Expanded capabilities for Bayesian analysis including the addition of a RANDOMstatement and multivariate priors in MCMC Production SURVEYPHREG procedure New model diagnostics in NLIN GRAPH license no longer required for ODS Graphics28These materials are proprietary to SAS and constitute SAS confidential information pursuant to the terms of the confidentiality or nondisclosure agreement in place between you and SASCopyright 2011, SAS Institute Inc. All rights reserved. 29. SAS/ETS 9.3 Forecasting and Econometrics Key Features New procedure for fitting copula distributions New procedure for statespace modeling Automatic variable selection for COUNTREG procedure Update data engines to support latest releases Additional / improved ODS Graphics plots29These materials are proprietary to SAS and constitute SAS confidential information pursuant to the terms of the confidentiality or nondisclosure agreement in place between you and SASCopyright 2011, SAS Institute Inc. All rights reserved. 30. SAS/OR 9.3 Operations Research Key Features - Optimization New network simplex linear programming algorithm Crossover feature for interior-point linear programming solver (experimental) Improvements to nonlinear solvers and introduction of a multi-start method for nonlinear programming Performance improvements for mixed integer programming solver Key Features - Simulation Using data to drive models empirical distributions, nonhomogeneous Poisson processes Improved usability to simplify navigation of large models Support for Windows 64 and remote SAS servers Automatic input distribution fitting using JMP for Simulation Studio30These materials are proprietary to SAS and constitute SAS confidential information pursuant to the terms of the confidentiality or nondisclosure agreement in place between you and SASCopyright 2011, SAS Institute Inc. All rights reserved. 31. Introducing SAS High-Performance Analytics SAS High Performance Analytics has been developed tosolve complex business problems that requiresophisticated high-end analytics and access to big data Statistical techniques addressed by the product include: Correlation analysis, Linear regression and variable selection with classical and modern methods, Logistic regression, Neural nets, Linear mixed models, Nonlinear regression and maximum likelihood, Optimization31Copyright 2011, SAS Institute Inc. All rights reserved. 32. SAS High-Performance Analytics: Topology(DRAFT)Base SAS1Client Tier SAS/STAT1SAS High Performance Analytics1SAS/Access Interface for Teradata orGreenplum1SAS Grid Manager2ApplianceSAS In-Memory AnalyticsSAS Embedded ProcessDatabase Storage Teradata or EMC Greenplum1 Dependent products, 2 Optional products32 Copyright 2011, SAS Institute Inc. All rights reserved. 33. SAS High-Performance Analytics:CapabilitiesData Preparation Data ExplorationAnalytics (HP)DS2* SUMMARY/MEANS HPLOGISTIC SORT FREQ HPREG RANK HPLMIXED HPFOREST HPNEURAL HPREDUCE HPNLIN 33 Copyright 2011, SAS Institute Inc. All rights reserved. 34. SAS High-Performance Analytics:ArchitectureSubmit a program from a SAS client session (eg. HPLOGISTIC)proc hplogistic data=GPlib.sgf_binary; class A B C; model y = a b c x1 x2 x3; performance details host="green1";run;34Copyright 2011, SAS Institute Inc. All rights reserved. 35. SAS High-Performance Analytics:Architecture Master Request is sent to the appliance and received by the Master NodeWorker Node 1Worker Node 2Worker Node N35 Copyright 2011, SAS Institute Inc. All rights reserved. 36. SAS High-Performance Analytics:Architecture Master Worker Node 1Worker Node 2 Worker Node N Analytical Computation and data request sent to the worker nodes36 Copyright 2011, SAS Institute Inc. All rights reserved. 37. SAS High-Performance Analytics:ArchitectureMasterWorker Node 1Worker Node 2 Worker Node N Data request sent to the database, data slice moved into memory 37Copyright 2011, SAS Institute Inc. All rights reserved. 38. SAS High-Performance Analytics:ArchitectureMasterWorker Node 1Worker Node 2 Worker Node NAnalytic Processing with internode communication 38Copyright 2011, SAS Institute Inc. All rights reserved. 39. SAS High-Performance Analytics:ArchitectureMasterWorker Node 1Worker Node 2 Worker Node NWorker node results returned to the Master Node, finalizecomputation 39Copyright 2011, SAS Institute Inc. All rights reserved. 40. SAS High-Performance Analytics:ArchitectureRoot NodeWorker Node 1Worker Node 2Worker Node NResult returned to the client40Copyright 2011, SAS Institute Inc. All rights reserved. 41. Much More Being Introduced41Copyright 2011, SAS Institute Inc. All rights reserved. 42. Call to Action to SAS Jedi Warriors Speak to your SAS Representative on TechnologyUpdates Join the SAS Singapore User Community SASCommunity.org May the Proc be with you! 42 Copyright 2011, SAS Institute Inc. All rights reserved. 43. Copyright 2011, SAS Institute Inc. All rights reserved.