Performance Assessment

⚠️ NOTE: This page describes a new model assessment and tuning workflow in the development version of mixOmics. We invite you to try it and share feedback during beta testing! To install, use:
devtools::install_github("mixOmicsTeam/mixOmics", ref = "6.31.4")
If you’re using the stable Bioconductor version, this page won’t apply.

The perf.assess() function in mixOmics is used to assess the performance of final supervised models after tuning and training, including models like (s)PLS-DA, (s)PLS, block (s)PLS-DA (DIABLO), and MINT (s)PLS-DA. It evaluates the model’s performance using cross-validation and outputs various metrics such as misclassification error rate, Balanced Error Rate (BER), Mean Squared Error of Prediction (MSEP), AUC, and ROC curves, depending on the model type. This page illustrates the use of perf.assess() with classification examples and regression examples. The predict() function can be used to mak predictions using a model on unseen data. Performance metrics help evaluate model accuracy and guide model improvements.

Data used on this page:
srcbt
liver.toxicity
linnerud

Key functions used on this page:
plsda()
pls()
perf.assess()
auroc()
predict()