How to calculate the hit rate of ks curve of logistic regression model?

The hit rate can be understood as the proportion of samples predicted by the model as positive examples, which is actually positive examples. In the ks curve of logistic regression model, the hit rate corresponds to the ordinate on the curve.

To calculate the hit rate, it is necessary to set a threshold first. The samples with the model prediction score greater than or equal to this threshold are regarded as positive examples, and the samples with less than this threshold are regarded as counter examples. Then, according to the comparison result between the actual tag and the predicted tag, the number of samples predicted as positive cases is counted and divided by the total number of samples predicted as positive cases, and the hit rate can be obtained.

For example, if the threshold is set to 0.5 and 100 samples are predicted as positive cases, and 80 of them are actually positive cases, the hit rate is 80%.

It should be noted that the hit rate is only one aspect of the sample prediction effect and cannot be used as an evaluation index alone. In practical application, the performance of the model can be comprehensively evaluated by combining other indicators such as accuracy, recall and F 1 value.

In a word, when calculating the hit rate in the ks curve of logistic regression model, it is necessary to set a threshold to count the proportion of positive cases in the samples predicted as positive cases. In the model evaluation, it is necessary to combine other indicators for comprehensive evaluation.