Skip to content

Metrics

Basic terms

  • Precision = True Positives / (True Positives + False Positives)
    • For all positive predicted, how many of them are correct.
    • i.e. How many hit you missed
  • Recall = True Positives / (True Positives + False Negatives)
    • How many positive is predicted among ALL ground true positive
    • i.e. How many REAL target you hit

Mean Absolute Error(MAE) :

To put it in short, take the absolute value of below error

error=actualpredicted

Comments