Getting Started

7 of the Most Used Regression Algorithms and How to Choose the Right One

Linear and Polynomial Regression, RANSAC, Decision Tree, Random Forest, Gaussian Process and Support Vector Regression

Regression Algorithms — Image by the author

Regression is a subset of Supervised Learning. It learns a model based on a training dataset to make predictions about unknown or future data. The description ‘supervised’ comes from the fact that the target output value is already defined and part of the training data. The difference between the subcategories Regression and Classification is only due to the output value. While Classification divides the dataset into classes, Regression is used to output continuous values. [Ras16]

Overview of types of learning — Image by the author

This article introduces a few of the most used Regression methods, explains some metrics to evaluate the performance of the models and describes how the model building process works.

• Regression methods
  - Multiple Linear Regression
  - Polynomial Regression
  - Robust Regression — RANSAC
  - Decision Tree
  - Random Forest
  - Gaussian process regression
  - Support Vector Regression