What is regression?

This article explains the concept of Regression as a Machine Learning term, and as a concept within the AI & Analytics Engine: Regression is a type of problem where one needs to predict a numeric value of an object

What type of problem is Regression? 

Regression is a type of problem where one needs to predict a numeric value of an object, for example:

• Predict how many units of products a customer will buy next week/month.

• Predict the total claim amount a health insurance customer will make over the next year, based on their lifestyle and medical history.

• Predict the total amount of monetary value a customer would spend during their entire time using the products that a business has to offer.

Proper data-driven prediction can assist in making faster and more accurate business decisions. Automating the process using machine-learning models is crucial, especially when dealing with a large amount of data.

Regression using the AI & Analytics Engine

State-of-the-art algorithms for regression modeling are available on the AI & Analytics Engine under the “Create App” function, with the option "Predict a variable/column" followed by choosing "Predict a numeric value".

The Engine automatically recommends this option when the data in the column selected for prediction (i.e., the “target column”) is numeric.

For further information about regression models, check out:

• Which metrics are used to evaluate a regression model's performance? 
• What is Prediction Quality and how is it calculated for the different ML problem types?