Model Evaluation & Metrics
Beginner
Classification Accuracy
Beginner
~8 min
code completion
Accuracy Score
Accuracy is the simplest classification metric — the fraction of predictions that match the ground truth:
In NumPy:
np.mean(y_true == y_pred)
The == produces a boolean array; mean converts True→1, False→0 and averages.
Limitation: accuracy is misleading on imbalanced datasets (e.g., 99% negative class — predicting all-negative gives 99% accuracy but zero useful signal).
Your task:
Implement accuracy(y_true, y_pred) that returns the fraction of correct predictions.
Example Tests
3 correct out of 5
Input: {"y_pred":[1,0,0,1,1],"y_true":[1,0,1,1,0]}
Expected: 0.6
All correct: accuracy = 1.0
Input: {"y_pred":[1,1,0,0],"y_true":[1,1,0,0]}
Expected: 1
All wrong: accuracy = 0.0
Input: {"y_pred":[1,1,1],"y_true":[0,0,0]}
Expected: 0