Discriminant Q(2) (DQ(2)) for improved discrimination in PLSDA models

In this paper we introduce discriminant Q(2) (DQ(2)) as an improvement for the Q(2) value used in the validation of PLSDA models. DQ(2) does not penalize class predictions beyond the class label value. With rigorous Monte Carlo simulations we show that when DQ(2) is used, a smaller effect can be found statistically significant than when the standard Q(2) is used.