This article provides a brief definition of the multinomial-logistic regression classification algorithm and its uses and benefits.

###

What is the Multinomial-Logistic Regression Classification Algorithm?

Logistic regression measures the relationship between the categorical target variable and one or more independent variables It deals with situations in which the outcome for a target variable can have two or more possible types. Logistic regression makes use of one or more predictor variables that can be either continuous or categorical and predicts the target variable classes. Logistic regression model output is helpful in identifying important factors that will impact the target variable and the nature of relationships between each of these factors and dependent variables.

This analysis reveals the following:

• Age - Multinomial logit (Natural log of the proportion of High Satisfaction to that of Medium satisfaction) estimate for 1 year increase in age for high job satisfaction relative to medium job satisfaction when other independent variables are held constant = 1.54
• Male vs. Female - Multinomial logit estimate for comparing male to females for high job satisfaction relative to medium job satisfaction when other variables are held constant = 0.67

How Does One Use the Multinomial-Logistic Regression Classification Algorithm?

Multinomial-Logistic Regression Classification can be applied to analyze numerous factors.

• Medical Diagnosis - Given a list of symptoms, one can predict if a patient is likely to be diagnosed with initial/intermediate/serious stages of a particular disease.
• Weather Prediction - Based on temperature, humidity, pressure etc. this analysis can predict rainy/sunny/cold weather.
• Satisfaction Analysis - Based on the attributes of a respondent e.g., demographics, marital status, gender, income, age, qualification etc., analysis can check the level of likely satisfaction with life/job/product/services.

Let’s look at two use cases:

Use Case – 1

Business Problem: A research agency wants to predict the likelihood of each election candidate being voted on by each voter and in turn devise a strategy to take proactive steps. The analysis can include specific data points such as ‘preferred party name’ and predictors might include customer demographics such as age, income, qualification, occupation, gender, religion and past voting status etc.

Business Benefit: By having knowledge of the probable election outcome, the organization can develop a proper strategy to address the discrepancies between expectations and predictions and identify the segments with a high likelihood of voting oppositions to effectively target voters and achieve more votes in favor of a particular candidate.

Use Case - 2

Business Problem: A doctor wants to predict the likelihood of a new patient having a disease that is in the initial/moderate/severe stage based on various health and body attributes of a patient such as blood pressure, hemoglobin, blood sugar, red blood count, etc.

Business Benefit: Given the profile of a patient and predicted level of disease, the doctor can determine the right treatment and/or medication.

The Multinomial-Logistic Regression Classification Algorithm is useful in identifying the relationships of various attributes, characteristics and other variables to a particular outcome.