I recently built a simple linear model that I trained using a standard 30-70 split on my data set. To my surprise, when I tested my model on unseen data, it reported the following:

enter image description here

With a linear score of 0.99999+.

It can't be overfitting as the model has never seen this data before, can it? What should I be wary of when using this model?


closed as unclear what you're asking by Stephan Kolassa, Peter Flom Aug 24 at 10:41

Please clarify your specific problem or add additional details to highlight exactly what you need. As it's currently written, it’s hard to tell exactly what you're asking. See the How to Ask page for help clarifying this question. If this question can be reworded to fit the rules in the help center, please edit the question.

  • 2
    $\begingroup$ There's most likely a conceptual relationship between the features you use to predict excess costs. Perhaps excess cost is the sum of many quantities and you're using all of them as inputs. Or they include measures of profit and revenue, or something along those lines. $\endgroup$ – CloseToC Aug 24 at 9:22
  • 1
    $\begingroup$ What @CloseToC says. Without knowing much more about your data, there is very little else we can tell you. $\endgroup$ – Stephan Kolassa Aug 24 at 9:29
  • $\begingroup$ I agree with the two comment so far. But, in general, yes, it certainly can be overfitting even on test data. Whether it is overfitting - well, we can't tell. $\endgroup$ – Peter Flom Aug 24 at 10:41