0

the dataset seems like non-linear, is there any recommended way to fit the datatset? since it's a non-linear regression problem, what's the correct way to evaluate the model's prediction? is the MSE score useful for this problems? enter image description here

Wuuu
  • 1
  • For evaluating predictions, this thread may be helpful. – Stephan Kolassa Feb 05 '24 at 12:19
  • Hi Stephan, you mention "If you want to predict the conditional expectation, use the MSE." Can I interpret it this way, if you want to predict what's the mean value of the target when inv = 2 , use MSE – Wuuu Feb 05 '24 at 13:28
  • Yes, exactly. The paper I referenced in that answer may be helpful; feel free to reach out to me on ResearchGate for it if you are interested. – Stephan Kolassa Feb 05 '24 at 14:38
  • thanks Stephan. one more question, as you can see, the target in this example is norm distributed and the mean is around 0, does the MSE still a good evaluator ? Intuitively, if we just use 0 as prediction, we will got the minimal MSE value. – Wuuu Feb 26 '24 at 01:53

0 Answers0