What does a negative q2 value mean in PLS-DA/OPLS-DA two-dimensional plots?
In a two-dimensional plot of PLS-DA (Partial Least Squares Discriminant Analysis) or OPLS-DA (Orthogonal Partial Least Squares Discriminant Analysis), the q² value (Q-squared) measures the cross-validation predictive effectiveness of the model on the data. Ideally, the q² value should be close to 1. A negative q² value indicates poor predictive power of the model, usually suggesting that the model has failed to effectively capture the relevant variation in the data or there is overfitting, indicating that the model's explanation or prediction of the data is not representative or accurate. In practice, if this situation arises, it is considered necessary to re-evaluate the model selection, parameter settings, or the accuracy of the data.
Biotyper - A leading service provider in bioproduct characterization and multi-omics mass spectrometry analysis.
Related services:
How to order?
