MBA 597: Special Topics in MIS
Business Analytics
Module 5
Model performance, evidence, and probabilities
Patrick Maher
pmaher@pointpark.edu
SAS Drive to start
Build Models
New Project
Name the project
Select the hospital LOS data
Click OK
Click Save
Assign a Target variable
Reject LENGTH_OF_STAY
Why? Collinearity
n statistics, correlation between predictor variables (or independent variables), such
that they express a linear relationship in a regression model. When predictor
variables in the same regression model are correlated, they cannot independently
predict the value of the dependent variable.
(ttps://www.britannica.com/topic/collinearity-statistics
Go to the Pipelines tab
Add a Decision Tree
As the name hints, letàadd a Logistic Regression
You should have two models
Click, Run over Model Comparison
If the models run successfully
Click, Results
You will see a champion model
Click the Assessment tab at top left
Click the dropdown for Lift Reports, select Cumulative
Captured Response %
Select the ROC dropdown to the left to get a ROC curve
to the right
MBA 594: Special Topics in Business Analytics
Machine Learning
Module 6
Model performance
Patrick Maher
pmaher@pointpark.edu
SAS Drive -> Explore and Visualize Data
Start with Data
Select the Heart data set
Hereàa view of the data
Select the Linear Regression object
Drag and drop the Linear Regression object
Use Diastolic as our Response variable, select Cholesterol as a Continuous effect, add Weight as a
Continuous effect,
Thereàa drop-down to select assessment criteria
Save as Heart linear regression
Name the regression and save
Build Models
New Project and name the project
Select Heart data
Save
Diastolic as Target
Select Linear Regression
Run from the regression node
Results from linear regression
Letàexamine the results

Purchase answer to see full
attachment