0
$\begingroup$

I use KNN regression to train out a model. The model estimates running time of a program based on different inputs, and the output is a single variable, which is time (double type).

I want to evaluate this model. I have a list of observation times and a list of prediction times. I would like to know how I can evaluate the model based on these two lists. What kind of measurements can I use?

Two lists for example:

Observation List: {0.01, 0.03, 0.02, 0, 0.02}

Prediction List: {0.05, 0.02, 0.04, 0.37, 0.05}

$\endgroup$
0
$\begingroup$

You can always use MSE/RMSE or even AIC/BIC.

$\endgroup$
  • $\begingroup$ I would like to know What kind of value of MSE/RMSE can be viewed as good prediction? $\endgroup$ – user3617042 May 8 '14 at 19:05

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.