Logistic Regression Models

Fit Regression Models for Nominal or Ordinal Responses

When your response variable has discrete values, you can use the Fit Model platform to fit a logistic regression model. The Fit Model platform provides two personalities for fitting logistic regression models. The personality that you use depends on the modeling type (Nominal or Ordinal) of your response column.

For nominal response variables, the Nominal Logistic personality fits a linear model to a multi-level logistic response function.

For ordinal response variables, the Ordinal Logistic personality fits the cumulative response probabilities to the logistic distribution function of a linear model.

Both personalities provide likelihood ratio tests for the model, a confusion matrix, and ROC and lift curves. When the response is binary, the Nominal Logistic personality also provides odds ratios (with corresponding confidence intervals).

Figure 11.1 Logistic Plot for a Nominal Logistic Regression Model 

Contents

Overview of the Nominal and Ordinal Logistic Personalities

Nominal Logistic Regression

Ordinal Logistic Regression

Other JMP Platforms That Fit Logistic Regression Models

Examples of Logistic Regression

Example of Nominal Logistic Regression

Example of Ordinal Logistic Regression

Launch the Nominal and Ordinal Logistic Personalities

Validation in Logistic Regression Models

The Logistic Fit Report

Whole Model Test

Fit Details

Lack of Fit Test

Logistic Fit Platform Options

Options for Nominal and Ordinal Fits

Options for Nominal Fits

Options for Ordinal Fits

Additional Examples of Logistic Regression

Example of Inverse Prediction

Example of Using Effect Summary for a Nominal Logistic Model

Example of a Quadratic Ordinal Logistic Model

Example of Stacking Counts in Multiple Columns

Statistical Details for the Nominal and Ordinal Logistic Personalities

Logistic Regression Model

Odds Ratios

Relationship of Statistical Tests