Root Mean Squared Error (RMSE)

#evaluation #machine-learning #deep-learning #interview
Root Mean Squared Error Formula

RMSE=1N∑i(xi−xi^)2

Pros

  1. Easy to understand
  2. Computationally simple
  3. Easy to differentiate
  4. Does not penalize the error as much as Mean Squared Error (MSE)

Cons

  1. Bad at Handling Outliers
  2. RMSE increases with the size of the test dataset

Related Notes