Reply to topic

LevenbergMarquardt for weighted nonlinear least squares

luoyubo
Registered Member
Posts
2
Karma
0
I am using Eigen's LM algorithm to fit a surface, but there is a lot of noise in the data, so I need to filter out the noise by iterative weighting.
But I don't know where this weight should be added?
you know, in my LevenbergMarquardt functor, I must fill fvec with:

for(i...)
fvec(i) = f_i(x)

But in the actual problem, each observation has a different weight, so how should I add the weight?
Can you help me with an example?
or did I post the question on the appropriate forum?
Thanks.

Reply to topic

Bookmarks



Who is online

Registered users: Baidu [Spider], Bing [Bot], cylverbak, funkyskywalker, Google [Bot], ineuw, Majestic-12 [Bot], maxmu, MRK0017, Sogou [Bot], sylvainrousseau, Yahoo [Bot]