Concordance probability method

Calculate the optimal probability classification threshold using the concordance probability method (CZ).

Usage

calc_cz(estimates, truth)

Arguments

estimates

A vector of classification probabilities. Values should represent the probability of 1 in the truth argument.

truth

An integer vector of 0 and 1 representing the true classifications.

Value

A numeric scalar representing the optimal probability threshold.

Details

The concordance probability method (Liu, 2012) is defined as the product of the sensitivity and specificity at a given threshold cut-point. By utilizing the product of both metrics, the value for the cut-point remains within the range of [0, 1]. The optimal threshold is the cut-point that maximizes the area of the rectangle related to the ROC curve.

The optimality criterion is then defined as:

$$\text{max}(sensitivity * specificity)$$

References

Liu, X. (2012). Classification accuracy and cut point selection. Statistics in Medicine, 31(23), 2676-2686. doi:10.1002/sim.4509

See also

Other optimal threshold methods: calc_gmean(), calc_topleft(), calc_youden()

Examples

calc_cz(estimates = dcm_probs$att1$estimate,
            truth = dcm_probs$att1$truth)
#> [1] 0.3170266

calc_cz(estimates = dcm_probs$att2$estimate,
            truth = dcm_probs$att2$truth)
#> [1] 0.4363361

calc_cz(estimates = dcm_probs$att3$estimate,
            truth = dcm_probs$att3$truth)
#> [1] 0.3722605