Algorithm 2 Personalized Classifier with Multiple Thresholds (PCMT).

Require:$X\_train$, $X\_test$, a set of $pcc\_thresholds$  Get Personalized Classifiers for each of the $pcc\_threshold$ in the $pcc\_thresholds$ for $X\_train$ and $X\_test$ (i.e., Personalized Classifier($X\_train,X\_test,pcc\_threshold$))  Obtain predicted probabilities from each of the PC classifiers for $X\_test$  Average the predicted probabilities of the PC classifiers  Return predicted probability for the $X\_test$