Senior Experience Machine Learning Section 6 - Polynomial Linear Regression

Polynomial Linear Regression

This one is similar to Multiple Linear Regression, but it uses exponents

y = b₀ + b₁x₁ + b₂x₂+… b_n*x_n

Polynomial Regression:

y = b₀ + b₁x₁ + b₂x₂² +… b_n*x_nⁿ

By using exponents, it can solve a bit more of the non-random cases, for example

It’s still called “Linear” Regression because the word linear is referring to the coefficients, not the predictor variables. The coefficients are still linear for now.

Here is the regressor applied to the Leaf Dataset:

Using a much more appropriate dataset that follows the assumptions with the same code:

Trying wacky data and seeing how it reacts:

Tweet Post Link