5.14.3. Multiclass Logistic Regression (Python clip0250 action)

<< Click to Display Table of Contents >>

Navigation:  5. Detailed description of the Actions > 5.14. Python Various >

5.14.3. Multiclass Logistic Regression (Python clip0250 action)

 

Icon: ANATEL~4_img119

 

Property window:

 

ANATEL~4_img118

 

Short description:

Creates a multinomial
logit predictive model

 

Long Description:

When you have a large number of variables (a few hundreds), all R libraries dealing with MNL modeling tend to fall apart as they first discretize every variable and need to run a full model before step-wise approaches are used.

 

In this very particular case, Python’s library is better implemented. This node will fit the model.