Optimization of Gaussian maximization of Likelihood

Hi,

I have a equation :

-0.5 x'* inv (K) * x - 0.5 log mod (K)

Have to find the parameters (s,l) for K that maximizes the above equation. where K = s * exp (|| x- x_0 ^2 / (2*l)) where x and x_0 is length of 500 samples

How to approach it ? Reference Paper: Rivet B, Niknazar M, Jutten C. Nonparametric modelling of ECG: applications to denoising and to single sensor fetal ECG extraction. In International Conference on Latent Variable Analysis and Signal Separation 2012 Mar 12 (pp. 470-477). Springer Berlin Heidelberg.

[Image Analyst]

What is mod(k)? Is that MATLAB's mod() function? The mod() function requires two arguments. Using it would make your signal non-linear. Do you have a snapshot of the actual formula you could post?

Have you looked at functions like fitglm() in the Statistics and Machine Learning Toolbox? It can get you best estimates for parameters in equations similar to that.