Uncertainty for regressed parameters using nlinfit considering the uncertainty of data
3 views (last 30 days)
Show older comments
Hello everyone,
I have some experimental data and I used a non-linear model to describe the data. The model has two independent variables (z-position and t-time) and one dependent variable . So, my experimental data is defined as , where each pair of value (z, t) leads to a unique value of . The model can be described as:
where is an empirical parameter, is the last time instance, and function f is defined by:
H, b are constants for a given problem, and I considered only the first 100 terms in n.
I applied the nlinfit algorithm, and I get the three estimated parameters of the model ( and D). Also, from CovB matrix I get the standard deviation (relative uncertainty) of these parameters.
Now, I would like to consider the uncertainty of my experimental data (which is known) and calculate its contribution in the total uncertainty of the parameters. How could I do this?
Thanks in advance.
0 Comments
Answers (0)
See Also
Categories
Find more on Regression in Help Center and File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!