The Matthews Correlation Coefficient (MCC) is used as a performance metric. It may be viewed as a discretization of the Pearson correlation for binary variables.
Parameters:
(1) true positive (TP), as a decimal fraction from 0 to 1
(2) true negative (TN), as a decimal fraction from 0 to 1
(3) false positive (FP), as a decimal fraction from 0 to 1
(4) false negative (FN), as a decimal fraction from 0 to 1
A =
= ((TP) * (TN)) - ((FP) * (FN))
B =
= (TP + FP) * (TP + FN) * (TN + FP) * (TN + FN)
Matthew's correlation coefficient = MCC =
= A / B
Interpretation:
• The closer the MCC is to 1 the better the performance.
To read more or access our algorithms and calculators, please log in or register.