2

I have a regression problem where I am predicting a continuous variable. Loss functions used most often in these cases (RMSE, MAE, etc.) don't treat over- or under- predictions differently.

I am in a scenario where under-predicting would be a much worse outcome than over-predicting.

What type of loss function would appropriately capture this?

user1566200
  • 315
  • 1
  • 3
  • 8
  • 2
    Use something like RMSE but multiply by alpha > 1 if it's an under prediction and by 0 < beta < 1 if it's an over prediction? – kbrose Sep 08 '17 at 14:06
  • Can the business cost of failed predictions be made explicit in your case? You might be able to make your business cost per example (or a transformation of it) the loss function, instead of a test metric. – Neil Slater Sep 08 '17 at 16:22
  • I'm not sure what 'under-' and 'over-' predictions are, but if you mean that your data is unbalanced towards a certain class, I suggest using Infogain Loss matrix to counter that – Alex Sep 12 '17 at 17:10

1 Answers1

1

Pick an asymmetric loss function. One option is quantile regression (linear but with different slopes for positive and negative errors).

Brian Spiering
  • 21,136
  • 2
  • 26
  • 109