What is multiclass classification?

This article explains the concept of Multi-class Classification as a Machine Learning term and as a concept within the AI & Analytics Engine: Multi-class Classification ar objects classified into one among a predetermined set of classes

Multi-class Classification 

Multi-class Classification is a type of Classification task where objects are classified into one among a predetermined set of classes. The number of classes can be two or more.

When there are exactly two classes and one is of particular interest, it is best to handle it as a Binary Classification task rather than a Multi-class Classification task.

Multi-class Classification occurs in many use cases, for example:

• Fruit classification and grading (superior-quality class, class I, or class II), are based on properties such as size, shape, color, weight, etc.

• Predicting what type of failure a machine is likely to undergo (power failure, overstrain, heat, or random failure), in predictive maintenance.

• Categorizing the sentiment of customer product reviews (positive, negative, or neutral).

Machine Learning (ML) models help to automate such classification processes to reduce manual labour, in order to satisfy businesses' requirements for large volumes of data prediction every day.

Multi-Class Classification in the AI & Analytics Engine 

In the AI & Analytics Engine, you can build multiclass classification models in the the app builder pipeline, where you’ll need to upload your data and choose the column to predict. If the chosen column has multiple classes, the Engine will automatically select multiclass classification as the only suitable prediction type.