- 10

- 0

## Main Question or Discussion Point

Are there any elaborated theory or method how to fit parameters of a function family to data given by probability distributions of data points instead of given coordinates of points precisely without error? I think this is a very general problem, I hope it is already solved.

Important:

I would like a general method working with any kind of probability distribution around data points, not just a Gaussian which can be described an error value, for example its variance.

I would like to use all information which is available, so a fully Bayesian solution without unnecessary estimation.

Important:

I would like a general method working with any kind of probability distribution around data points, not just a Gaussian which can be described an error value, for example its variance.

I would like to use all information which is available, so a fully Bayesian solution without unnecessary estimation.