Machine LearningLesson 77Free preview

Polynomial Regression

This lesson is part of Python · All levels

Summary

Polynomial regression models curved relationships by adding the rank of variables such as x squared and x rank three as additional features. In scissor-learn, PolynomialFeatures convert one column X into a matrix containing its ranks that are then assigned to LinearRegression. The higher the degree, the more flexible the lines, but the risk of overfitting also increases. A degree of 2 or 3 is usually sufficient for a simple curved relationship.

Key points

Polynomial regression deals with nonlinear relationships between variables
PolynomialFeatures make derivative features such as x^2 and x^3
High degrees make the line more flexible but prone to overfitting
The transformation resulting features are trained with LinearRegression

Full lesson

Code examples and quizzes

Full lessons, code examples, and quizzes are available to subscribers. 5 questions.