I'm looking to do polynomial regression on 2 relatively uncorrelated features. Even with very high degrees, I find that the model doesn't fit very well to the data (figure attached here). Since the number of degrees is the only parameter that I can modify, I am looking to know if polynomial regression only works with highly correlated variables?

enter image description here

  • $\begingroup$ What do the various colors represent? Linear regression assumes that there is a linear relation between the variables (similar for polynomial). If this is not the case then you have the wrong model. $\endgroup$ Sep 20, 2019 at 6:14
  • 1
    $\begingroup$ This is a nice example of why polynomial regression can be a horrible regression procedure: the right tail of the fit is determined primarily by the data at the extreme left. It's obviously a terrible description of the right-hand points. It's likely you will learn far more by choosing effective plots of the data and looking at them closely before continuing: consider using logarithmic or square root scales on both axes. $\endgroup$
    – whuber
    Sep 20, 2019 at 14:26


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Browse other questions tagged or ask your own question.