Let's consider a general signal processing estimation problem where the measurements are modeled as $${\bf x}[n]={\bf s(\theta)}+{\bf w}[n],$$ where ${\bf w}$ is a non-Gaussian r.v. (noise term) and ${\bf s(\theta)}$ is a deterministic nonlinear signal model. Since the noise is non-Gaussian I am wondering how one could design a Maximum Likelihood estimator for the parameter vector ${\bf \theta}$ if an estimate of the pdf of ${\bf w}$ was obtained by density estimation or another nonparametric technique. I have looked into empirical likelihood but I am not sure if it applies to this problem. Is this a standard problem in the context on non-parametric statistics? References are welcome.


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.