LEC7 - An Example of Building a Prediction Model Using the...

Info icon This preview shows pages 1–4. Sign up to view the full content.

View Full Document Right Arrow Icon
An Example of Building a Prediction Model Using the Logistic Regression Node SECTION 1 INTRODUCTION ................................................................................................... 3 SECTION 2 ILLUSTRATIVE EXAMPLE ............................................................................... 3 D ATA ........................................................................................................................................... 3 P URPOSE OF THE S TUDY ............................................................................................................... 4 C OMPLETE D IAGRAM ................................................................................................................... 4 S TEP 1: D ATA S OURCE N ODE ....................................................................................................... 4 S TEP 2: F IXING D ATA P ROBLEMS WITH SAS C ODE N ODE ........................................................... 5 S TEP 3: U SING THE I MPUTE N ODE TO I MPUTE V ARIABLES WITH S MALL A MOUNT OF M ISSING V ALUES ....................................................................................................................................... 8 S TEP 4: C REATING M ISSING V ALUE S I NDICATOR WITH SAS C ODE N ODE ................................. 10 S TEP 5: C REATING MVPVARIABLE WITH SAS C ODE N ODE .................................................. 11 S TEP 6: R EDUCING THE N UMBER OF L EVELS AND I MPUTING M ISSING V ALUES FOR C ATEGORICAL V ARIABLES WITH SAS C ODE N ODE ................................................................... 12 S TEP 7: I MPUTE N UMERICAL V ARIABLES WITH C LUSTER N ODE ................................................ 15 S TEP 8: D ROPPING U NWANTED V ARIABLES WITH D ROP N ODE .................................................. 16 S TEP 9: U PDATING M ODEL R OLE AND M EASUREMENT L EVEL WITH M ETADATA N ODE ............ 17 S TEP 10: F ITTING A P REDICT M ODEL WITH R EGRESSION N ODE ................................................ 19 S TEP 11: U NDERSTANDING THE RESULTS AND SELECTING THE CUT OFF PROBABILITY ............... 22 S TEP 12: P ERFORMING O VER S AMPLING WITH S AMPLE N ODE ................................................... 26 S TEP 13: R EFINING P REDICT M ODEL WITH R EGRESSION N ODE ................................................. 29 S TEP 14: U NDERSTANDING R EGRESSION R ESULTS .................................................................... 30 S TEP 15: C OMMUNICATING R ESULTS TO A G ENERAL A UDIENCE ............................................... 32 APPENDIX 1 SAS CODE FOR FIXING DATA PROBLEM ................................................ 38 APPENDIX 2 SAS CODE FOR MISSING VALUE INDICATOR ....................................... 39 APPENDIX 3 SAS CODE FOR MISSING VALUE PATTERN ........................................... 40 APPENDIX 4 SAS CODE FOR CATEGORICAL VARIABLE LEVEL REDUCTION AND MISSING VALUE IMPUTATION ................................................................................. 41
Image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
©Morgan C. Wang and Mark E. Johnson 2 Table 1 MVP Distribution after Grouping .................................................................. 12 Table 2 Style Distribution after Grouping ................................................................. 13 Table 3 ZIP3 Distribution after Grouping .................................................................. 13 Table 4 STATE Distribution after Grouping .............................................................. 14 Table 5 SOURCE Distribution after Grouping ........................................................... 14 Table 6 Model Fitted Statistics at Probability = 0.5 .................................................. 22 Table 7 Prior Probability ............................................................................................. 23 Table 8 Model Fitted Statistics at Probability = 0.057426 ........................................ 26 Table 9 Prior Probability ............................................................................................. 31 Table 10 Model Fitted Statistics at Probability = 0.057426 ...................................... 32 Figure 1 Cumulative Lift Chart ................................................................................... 24 Figure 2 Lift Chart ....................................................................................................... 25 Figure 3 ROC Curve .................................................................................................... 25 Figure 4 Grouped ZIP3 Performance ............................... Error! Bookmark not defined. Figure 5 Grouped State Performance .............................. Error! Bookmark not defined. Figure 6 Grouped Source Code Performance ................ Error! Bookmark not defined. Figure 7 Missing Value Pattern Performance ................. Error! Bookmark not defined. Figure 8 Triathlete Performance ...................................... Error! Bookmark not defined. Figure 9 Age Performance ................................................ Error! Bookmark not defined. Figure 10 Miles per Week Performance ........................... Error! Bookmark not defined. Figure 11 Races per Year Performance ........................... Error! Bookmark not defined.
Image of page 2
©Morgan C. Wang and Mark E. Johnson 3 Section 1 Introduction Using Enterprise Miner to build a moderately accurate prediction model is fairly straightforward. However, it is very challenging to build a very accurate prediction model using only those nodes available in Enterprise Miner since not all SAS functionalities are currently available in Enterprise Miner. For example, the logit plot is an excellent visualization tool for studying the relationship between a binary target and a set of continuous input variables. Unfortunately, the logit plot is not available in Enterprise Miner.
Image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 4
This is the end of the preview. Sign up to access the rest of the document.

{[ snackBarMessage ]}

What students are saying

  • Left Quote Icon

    As a current student on this bumpy collegiate pathway, I stumbled upon Course Hero, where I can find study resources for nearly all my courses, get online help from tutors 24/7, and even share my old projects, papers, and lecture notes with other students.

    Student Picture

    Kiran Temple University Fox School of Business ‘17, Course Hero Intern

  • Left Quote Icon

    I cannot even describe how much Course Hero helped me this summer. It’s truly become something I can always rely on and help me. In the end, I was not only able to survive summer classes, but I was able to thrive thanks to Course Hero.

    Student Picture

    Dana University of Pennsylvania ‘17, Course Hero Intern

  • Left Quote Icon

    The ability to access any university’s resources through Course Hero proved invaluable in my case. I was behind on Tulane coursework and actually used UCLA’s materials to help me move forward and get everything together on time.

    Student Picture

    Jill Tulane University ‘16, Course Hero Intern