Multinomial Logistic Regression
This section builds upon the concepts discussed in the Logistic Regression section. In this section we will focus on unordered choices.
Random Utility Basis
For the ith consumer faced with J choices, the utility of choice j is
Uij=w′iβj+z′ijγj+εij
Variables are defined as follows:
- Uij:= Utility an individual i gets while consuming product j.
- wi′:= Individual i′s characteristic like income, sex etc. Note that these characteristics do not vary with products.
- zij′:= Attributes of product j and some attributes can vary across individuals, eg. Transit time.
If the consumer chooses j, then
P[Uij>Uik],∀k=j.
Assume there are 4 choices {A,B,C,D}. Then the utilities for the ith consumer are given as
UiA