I have a non-linear regression function that outputs $f_1(2)=7\pm 0.1$ and say if I reduce the dimensionality of the parameters of the function and create second function that outputs $f_2(2)=6.7 \pm 0.8$ at the same input $2$.

Question: Is it possible to determine the relative information loss by only looking at the outputs of the two functions?

  • $\begingroup$ why your two functions are the same? $\endgroup$ Commented Apr 1, 2021 at 13:49
  • $\begingroup$ @GENIVI-LEARNER sorry, fixed now. $\endgroup$
    – user0193
    Commented Apr 1, 2021 at 13:50


Browse other questions tagged or ask your own question.