Polynomial Regression

📈 Polynomial Regression

Polynomial regression is an extension of linear regression where we model a nonlinear relationship between input $𝑥$ and output $𝑦$ by adding powers of $𝑥$.

Sometimes by defining a new feature you might get a better Model that requires less computation

Use case:

When prediction fits Polynomial equation instead of linear equation

For example house price can defined by calculating area instead of creating 2 variable equation we can define one variable equation :

• Scaling of feature becomes crucial in Polynomial Regression
• Some algo can choose feature to fit polynomial curves

Feature engineering

is the process of creating new features from existing ones to improve model performance.

Generic polynomial regression model of degree n:

$y = \theta_0 + \theta_1 x + \theta_2 x^2 + \theta_3 x^3 + ... + \theta_n x^n$

---

---

Related Posts

• Cost Function Regularization — how to tame the overfitting these higher-degree models are prone to
• Normal Equation in Linear Regression — a closed-form alternative to gradient descent for fitting these models