Matthews Correlation Coefficient (MCC) Calculator

Your details

Cases where your model predicted positive and the actual class is also positive. Also called hits.
Cases where your model predicted negative and the actual class is also negative. Also called correct rejections.
Cases where your model predicted positive but the actual class is negative. Also called false alarms or Type I errors.
Cases where your model predicted negative but the actual class is positive. Also called misses or Type II errors.
Matthews Correlation CoefficientStrong
0.7156

Ranges from -1 (worst) through 0 (random) to +1 (perfect)

Normalized MCC0.8578
Sensitivity / Recall90.9%
Specificity80%
Precision (PPV)83.3%
Accuracy85.7%
F1 Score0.8696
Total Samples105
0.7156
Inverse<-0.3Weak/negative-0.3-0Weak0-0.3Moderate0.3-0.5Good0.5-0.7Strong0.7+

MCC is 0.7156, indicating strong classification.

  • Overall accuracy is 85.7%, but MCC is a better summary on imbalanced data because it uses all four confusion-matrix cells.
  • Sensitivity and specificity are closely balanced (90.9% and 80.0%), so neither class is heavily favoured.
  • Your dataset has 52.4% actual positives out of 105 samples. MCC stays reliable even when this fraction is far from 50%.

Next stepThe normalized MCC is 0.8578, which maps the same information to the [0, 1] scale used by accuracy and F1 score, making it easier to compare across tools.

 = Powered by OnlyCalculators