0

Can I apply linear regression to any function to the x's. For example:

Polynomials with one variable: y = w_0 + w_1 * x + w_2 * x2 + ... + w_D * xD

Polynomials with multiple variables (let's say there are 2 variables and the order of the polynomial is also 2, so it's easy to write): y = w_0 + w_1 * x_1 + w_2 * x_2 + w_3 * x_1 * x_2 + w_4 * (x_1)2 + w_5 * (x_2)2

Radial basis functions: y = w_0 + w_1 * f(x_1) + w_2 * f(x_2) + ... + w_D * f(x_D), where f(x) is a radial basis function.

A89
  • 1
  • 1
  • Your idea is a good one and this question has a simple answer: yes. It is demonstrated by examining threads on this site about forms of non-linear regression. I have chosen a generic explanation as the duplicate, but would encourage you to look at some of the well-described, real-world examples you can find here. – whuber Mar 02 '19 at 13:29
  • 1
    Thank you, because I am digging more into this site, with a better idea about what answers could be helpful for me. – A89 Mar 02 '19 at 18:35

0 Answers0