Knowledge Maxxing← Back to topic library
ML MaxxingBeginner60 sec
Accuracy
Accuracy is the proportion of a classifier's predictions that are correct out of all predictions made.
Prerequisite: Classification
What matters
Accuracy can be misleading on imbalanced data, where a model can score very high accuracy simply by always predicting the majority class, without being useful at all.
Example
A model that predicts 'not fraud' for every transaction can be 99% accurate if only 1% of transactions are actually fraudulent, despite catching zero fraud.
Common misconception
High accuracy does not always mean a good model; on imbalanced data, other metrics like precision and recall are needed to see the full picture.
Recall question
Why can a model be 99% accurate and still be completely useless?
Keywords and related concepts
accuracyclassification metricimbalanced dataprecision