Detailed Confusion Matrix Analysis

Clear breakdown of correct vs incorrect predictions

Confusion Matrix
Predicted 0
Predicted 1
Actual 0
Correct
130
Incorrect
250
Actual 1
Incorrect
165
Correct
246
0
Class 0 Predictions
Correctly Predicted as 0
130
Incorrectly Predicted as 1
250
Class 0 Accuracy: 34.2%
130 correct out of 380 total
1
Class 1 Predictions
Incorrectly Predicted as 0
165
Correctly Predicted as 1
246
Class 1 Accuracy: 59.9%
246 correct out of 411 total

🚨 Key Finding: Performance Below Random

Overall Accuracy: 47.5% (376 correct / 791 total) - This is worse than random guessing which would achieve ~50%

Performance Metrics Summary

Overall Accuracy
47.5%
376 / 791 correct
Precision (Class 1)
49.6%
246 / 496 predicted as 1
Recall (Class 1)
59.9%
246 / 411 actual 1s
Specificity (Class 0)
34.2%
130 / 380 actual 0s
Total Errors
415
250 + 165 wrong predictions
Random Baseline
~50%
Expected random performance