Chapter 6: Classifying Trees with Multiclass Logistic Regression

Multiclass logistic regression is the Machine Learning (ML) algorithm used to classify events, entities, and behaviors into a fixed number of categories. It can be used across different industries and business scenarios when it's necessary to predict the classification of an entity into multiple groups. A typical classification use case is represented by the desire to segment the customer base of a company according to their profitability and preferences in order to target the right customers with the most effective marketing campaigns.

This kind of technique is an extension of the binary logistic regression that allows us to overcome the limits of two possible labels and opens the applicability to other contexts where we can find multiple categories to identify.

In this chapter, we'll see all the stages necessary to implement, evaluate, and test a multiclass logistic regression model leveraging BigQuery...