Use this metric to summarize one aspect of a classifier's confusion matrix after defining the positive class and counting outcomes on an appropriate evaluation set.

True positives

Required integer input.

True negatives

Required integer input.

False positives

Required integer input.

False negatives

Required integer input.

Calculate the mean of sensitivity and specificity. The tool evaluates the supplied inputs together and returns the named outputs below; it does not infer omitted operating conditions or change the units shown.^[1]

Balanced accuracy

The resulting balanced accuracy returned as a number.

• A single classification metric can conceal class imbalance, threshold tradeoffs, subgroup errors, calibration, and uncertainty. Results are only comparable when the positive class, averaging rule, dataset, and decision threshold match.
• Use finite inputs in the displayed units and preserve more precision than the final presentation requires. Independently verify safety-critical, financial, compliance, or production decisions.

Inspect the complete confusion matrix and report complementary precision, recall, specificity, balanced accuracy, F-score, calibration, and threshold curves as appropriate to the decision cost.

• Accuracy Calculator

  Calculate classification accuracy from true and false outcome counts.

• F1 Score Calculator

  Calculate the harmonic mean of classification precision and recall.

• Recall Calculator

  Calculate sensitivity from true positive and false negative counts.