How does Matlab calculate mean squared error in sbiofit?

Question

Oscar 2023-1-18

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1896515-how-does-matlab-calculate-mean-squared-error-in-sbiofit

评论： Oscar 2023-1-19

I'm fitting a simbiology model to experimental data. After the fit I can see that the model fits the data pretty well. However, the MSE property in the optimresults object returns Inf even though the residuals are not Inf (as shown from the optimresults Output property).

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Arthur Goldsipe 2023-1-18

1
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1896515-how-does-matlab-calculate-mean-squared-error-in-sbiofit#answer_1151540

The MSE could be Inf for several reasons. My guess is that you have a poorly posed fitting problem. Did you see any warning messages during the fit? That might give you additional clues.

Also, you can find the MATLAB code that does this calculation in function calcCovBeta of class SimBiology.fit.internal.FitObject, but let me summarize the calculations here. MSE is the sum of the sequares of the residuals divided by the degrees of freedom. If you try to estimate 2 parameters with 2 observations, your degrees of freedom will be 0, and the MSE will be Inf.

If you still don't understand why the MSE is Inf after reading these details, then I suggest sharing reproduction steps. Then, we can step through the calculations in the debugger and see exactly where and why the Inf value occurs.