Working point is an easy knob to use in order to optimize the model performance to fit our needs.
Working point (WP) is a value between 0 and 1 that represents the threshold we use in a binary classification model. Anything above this threshold will be classified as having label 1 (positive) and anything below this threshold will be classified as having label 0 (negative).
A commonly used working point is 0.5, however the working point is problem-dependent and can be changed in order to optimize the model performance to fit our needs.
The value of the working point will directly affect the confusion matrix and thus the false positive, false negative, true positive and true negative samples. Higher working point means less false positives and more false negatives and vice versa.
When optimizing model performance, we often look at the precision-recall curve. When we care most about identifying positive samples, we can lower the working point and "pay" with higher false positive rate. On the contrary, when we want to make sure that every positive prediction we make is correct, we can raise the working point and "pay" with higher false negative rate.
#explainableai #performance