Main Concept
Accuracy measures the percentage of total predictions that were correct out of all predictions made.
Formula (conceptual)
Accuracy = (Correct predictions) / (Total predictions)
When to Use
Balanced datasets where both classes appear with similar frequency.
Weakness
On imbalanced datasets, accuracy is misleading. A model that always predicts the majority class can achieve very high accuracy while being completely useless at detecting the minority class.
Exam Example
A dataset has 990 legitimate transactions and 10 fraudulent ones. A model that always predicts “not fraud” achieves 99% accuracy but detects zero fraud cases — making accuracy a poor metric for this scenario.
Exam Domain
- Domain 1, Task Statement 1.3: model performance metrics.
Related Notes
Links:
References