Notes: sensitivity & specificity

>**terminology**: *True positive* (TP); *False positive* (FP): originally negative; *True negative* (TN); *False negative* (FN): originally positive;

True positive rate (TPR, Sensitivity): TP / (TP + FN);
True negative rate (TNR, Specificity): TN / (TN + FP) = 1 - FPR;
False positive rate (FPR): FP / (FP + TN) = FP / N;
Accuracy (classifier) : (TP + TN) / (P + N).

Type 1 error: incorrectly reject the true value, increase as alpha increases;
Type 2 error: incorrectly accept the false value, increase as alpha decreases.