Enterprise Data Mining
with SQL Server



      Mark Tabladillo Ph.D.
      Microsoft MVP
      MarkTab Consulting http://marktab.com
      Blog http://marktab.net/datamining Twitter @marktabnet
      SQL Saturday 108 Redmond, WA
      February 25, 2012
About Mark Tabladillo

     20 Years in Atlanta, Georgia
     Consulting since 1998; Incorporated 2003
        Part-Time Faculty at University of Phoenix
     SAS and Microsoft Expert
        Presenter since 1998 at conferences like Microsoft
         TechEd and SAS Global Forum
     Taught statistics at undergraduate and graduate level
     Blog: http://marktab.net    @MarkTabNet



3
Enterprise:
Leaders of Leaders of
      Leaders
Enterprise Challenge
Enterprise Challenge
Enterprise Challenge
Enterprise Challenge
“Data Mining”
Definitions

Phrase          Goal
“Data Mining”   Inform actionable decisions


“Machine        Determine best performing
Learning”       algorithm
SQL Server
     2008 R2:

Physical and Logical
OLAP Engine
Physical
Architecture
 http://msdn.microso
  ft.com/en-
  us/library/ms17477
  6.aspx
Analysis Services
Logical Architecture
 http://msdn.microsoft.com/en-us/library/ms174587.aspx
Outline

 Contoso Retail and Fundamentals
 Enterprise-Level Data Mining Demo for
  SQL Server
 What is my next step?
What is Contoso Retail?

 Demonstration dataset for SQL Server
  Database Engine and Analysis Services
   http://www.microsoft.com/downloads/en/details.aspx?displaylang=en&FamilyID=868662dc-187a-
    4a85-b611-b7df7dc909fc
What are the fundamentals?


                          ‘Readin’

   Arithmetic   Reading
                          ‘Ritin’


           Writing        ‘Rithmetic
What Enterprise Tools support Data Mining?

 SQL Server Management Studio (SSMS)
 Business Intelligence Development Studio
  (BIDS)
    SQL Server Integration Services (SSIS)
 PowerShell version 2
What Enterprise Tools support Data Mining?




                  Data
                 Mining


  SSMS            SSIS        PowerShell
Variable      0   1   2   3   4   5   6   7



Discretized
Discretized
Continuous
Discrete
Variable      0   1   2   3   4   5   6   7



Discretized
Discretized
Continuous
Discrete
Variable      0   1   2   3   4   5   6   7



Discretized
Discretized
Continuous
Discrete
Variable      0   1   2   3   4   5   6   7



Discretized
Discretized
Continuous
Discrete
Variable      0   1   2   3   4   5   6   7


Discretized
Discretized
Continuous
Discrete
Documentation

 Data Mining Structures
   http://msdn.microsoft.com/en-
    us/library/cc645741.aspx
   http://msdn.microsoft.com/en-
    us/library/ms174757.aspx
 Data Mining Models
   http://msdn.microsoft.com/en-
    us/library/cc645779.aspx
Contoso Retail:
Enterprise Data Mining

   Demonstration
What is my next step?
 SQL Server 2008 R2 Enterprise
  (includes database engine, Analysis Services,
  SSMS and BIDS)
    http://www.microsoft.com/sqlserver/2008/en/us/trial-
     software.aspx
 Microsoft Office 2010 Professional
    http://office.microsoft.com/en-us/try
 PowerShell 2.0
    http://support.microsoft.com/kb/968929
 Data Mining Portal and Blog
    http://www.marktab.net
Conclusion

 Data mining leaders can tackle enterprise
  data mining challenges with
   SQL Server Management Studio
   Business Intelligence Development Studio
   PowerShell version 2
 Become leaders of leaders of leaders
Where Can I Find More Information?

 http://marktab.net Data Mining Resource
 http://marktab.net/datamining Data Mining
  Blog
 http://sqlserverdatamining.com SQL Server
  Data Mining
 http://technet.microsoft.com Microsoft’s
  TechNet
Graphics

 Ship graphics Copyright © 1995-2006 Nova
  Development and its licensors. All rights
  reserved. Used with permission.
Abstract
     This presentation introduces SQL Server Data Mining (SSDM) for SQL
     Server Professionals based on the speaker's past presentation for
     Microsoft TechEd. Starting with SQL Server Management Studio
     (SSMS), the demo includes the interfaces important for professional
     development, including Business Intelligence Development Studio
     (BIDS), highlighting Integration Services, and PowerShell. The
     interactive demos are based on Microsoft's Contoso Retail sample
     data. Finally we will evaluate where Microsoft data mining can help you
     in a practical business environment, which may include Oracle and
     SAS.




35

SQL Saturday 108 -- Enterprise Data Mining with SQL Server

  • 1.
    Enterprise Data Mining withSQL Server Mark Tabladillo Ph.D. Microsoft MVP MarkTab Consulting http://marktab.com Blog http://marktab.net/datamining Twitter @marktabnet SQL Saturday 108 Redmond, WA February 25, 2012
  • 3.
    About Mark Tabladillo  20 Years in Atlanta, Georgia  Consulting since 1998; Incorporated 2003  Part-Time Faculty at University of Phoenix  SAS and Microsoft Expert  Presenter since 1998 at conferences like Microsoft TechEd and SAS Global Forum  Taught statistics at undergraduate and graduate level  Blog: http://marktab.net @MarkTabNet 3
  • 4.
  • 5.
  • 6.
  • 7.
  • 8.
  • 9.
  • 10.
    Definitions Phrase Goal “Data Mining” Inform actionable decisions “Machine Determine best performing Learning” algorithm
  • 11.
    SQL Server 2008 R2: Physical and Logical
  • 12.
  • 13.
    Analysis Services Logical Architecture http://msdn.microsoft.com/en-us/library/ms174587.aspx
  • 14.
    Outline  Contoso Retailand Fundamentals  Enterprise-Level Data Mining Demo for SQL Server  What is my next step?
  • 15.
    What is ContosoRetail?  Demonstration dataset for SQL Server Database Engine and Analysis Services  http://www.microsoft.com/downloads/en/details.aspx?displaylang=en&FamilyID=868662dc-187a- 4a85-b611-b7df7dc909fc
  • 16.
    What are thefundamentals? ‘Readin’ Arithmetic Reading ‘Ritin’ Writing ‘Rithmetic
  • 17.
    What Enterprise Toolssupport Data Mining?  SQL Server Management Studio (SSMS)  Business Intelligence Development Studio (BIDS)  SQL Server Integration Services (SSIS)  PowerShell version 2
  • 18.
    What Enterprise Toolssupport Data Mining? Data Mining SSMS SSIS PowerShell
  • 21.
    Variable 0 1 2 3 4 5 6 7 Discretized Discretized Continuous Discrete
  • 22.
    Variable 0 1 2 3 4 5 6 7 Discretized Discretized Continuous Discrete
  • 23.
    Variable 0 1 2 3 4 5 6 7 Discretized Discretized Continuous Discrete
  • 24.
    Variable 0 1 2 3 4 5 6 7 Discretized Discretized Continuous Discrete
  • 25.
    Variable 0 1 2 3 4 5 6 7 Discretized Discretized Continuous Discrete
  • 26.
    Documentation  Data MiningStructures  http://msdn.microsoft.com/en- us/library/cc645741.aspx  http://msdn.microsoft.com/en- us/library/ms174757.aspx  Data Mining Models  http://msdn.microsoft.com/en- us/library/cc645779.aspx
  • 27.
    Contoso Retail: Enterprise DataMining Demonstration
  • 28.
    What is mynext step?  SQL Server 2008 R2 Enterprise (includes database engine, Analysis Services, SSMS and BIDS)  http://www.microsoft.com/sqlserver/2008/en/us/trial- software.aspx  Microsoft Office 2010 Professional  http://office.microsoft.com/en-us/try  PowerShell 2.0  http://support.microsoft.com/kb/968929  Data Mining Portal and Blog  http://www.marktab.net
  • 32.
    Conclusion  Data miningleaders can tackle enterprise data mining challenges with  SQL Server Management Studio  Business Intelligence Development Studio  PowerShell version 2  Become leaders of leaders of leaders
  • 33.
    Where Can IFind More Information?  http://marktab.net Data Mining Resource  http://marktab.net/datamining Data Mining Blog  http://sqlserverdatamining.com SQL Server Data Mining  http://technet.microsoft.com Microsoft’s TechNet
  • 34.
    Graphics  Ship graphicsCopyright © 1995-2006 Nova Development and its licensors. All rights reserved. Used with permission.
  • 35.
    Abstract This presentation introduces SQL Server Data Mining (SSDM) for SQL Server Professionals based on the speaker's past presentation for Microsoft TechEd. Starting with SQL Server Management Studio (SSMS), the demo includes the interfaces important for professional development, including Business Intelligence Development Studio (BIDS), highlighting Integration Services, and PowerShell. The interactive demos are based on Microsoft's Contoso Retail sample data. Finally we will evaluate where Microsoft data mining can help you in a practical business environment, which may include Oracle and SAS. 35