I was wondering whether it is possible to regularize (L1 or L2) non-linear parameters in a general regression model. Say, I have the following non-linear least squares cost function, where $p$ is a $3d$ vector of fitting parameters:

$cost(p_i) = ( y(x) - sin^{p_1}(x) + p{_2}e^{(p_3*x)} )^2$

In the above cost function, $p_1$ and $p_3$ are non-linear parameters. How should I go about regularizing them? If they were linear, I can just sum them up together with the linear parameters (absolute values or squares) and add as a penalty to the cost function, right? However, I'm not sure if I'm allowed to do so for non-linear parameters.

Has anyone considered this problem?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.