EasyDeepLearn

What is sMAPE and why is it used?

medium

Answer

  • Symmetric MAPE = mean(|y    yhaty\; - \;y_{\mathrm{hat}}| / ((y  +  yhat)/2)(( \mid y \mid \; + \; \mid y_{\mathrm{hat}} \mid ) / 2)) * 100%.
  • Bounded in [0%, 200%] and symmetric in over-/under-prediction: over-forecasting and under-forecasting by the same absolute amount give equal error.
  • Still undefined when both y and yhaty_{\mathrm{hat}} are 0.
  • Better than MAPE for time series forecasting competitions (used in M-competitions).
  • Not a proper scoring rule for probabilistic forecasts — MASE (mean absolute scaled error) is a better alternative in many cases.
Check yourself — multiple choice
  • sMAPE is always identical to MAPE
  • Symmetric MAPE — bounded in [0%, 200%], symmetric in over-/under-prediction
  • sMAPE is only for classification
  • sMAPE ignores the sign of the error

sMAPE symmetrizes MAPE and bounds the error — better for forecasting competitions.

#metrics-regression

Practise Supervised Learning

215 interview questions in this topic.

Related questions