Introduction to Predictive Modeling Using IBM SPSS Modeler (v18) eLearning

1: Introduction to predictive modeling

• Identify measurement levels
• Identify modeling objectives
• Identify three types of models
• Plan of the course

2: Examine individual fields

• Use measures of central tendency and dispersion
• Explain what are standardized scores (z-scores)
• Detect outliers in continuous fields
• Use automated data preparation

3: Examine the relationship between two fields

• Identify the steps in statistical testing
• Examine the relationship between two categorical fields using percentages and the Chi-Square test
• Examine the relationship between a categorical and continuous field using means and the F test
• Examine the relationship between two continuous fields using the correlation
• Identify important predictors with the Feature Selection node

4: Predictive modeling for a categorical target: CHAID

• Review the relationship between two categorical field
• Explain how CHAID builds a decision tree
• Assess the accuracy of a decision tree
• Score records with the CHAID model nugget to add propensities

5: Predictive modeling for a continuous target: Linear regression

• Review the relationship between two continuous fields.
• Explain how Linear regression predicts a continuous target
• Assess the accuracy of a Linear regression model
• Include categorical predictors in the model

6: Explore a clustering model to create customer segments

• Explain how K-Means creates segments
• Use the K-Means model nugget to score records

7: Overview of models

• Overview of classification models
• Overview of segmentation models
• Overview of association models