The self-validating ensemble models (SVEM) method is a resampling method where each row is given non-zero weights in a training and validation set. This method is particularly helpful in building models with sparse data sets or experimental designs where there are not enough observations to carry out traditional validation methods.
In the below video, we review the following:
- What is SVEM and how does it work?
- Using SVEM in the Gen Reg platform
- Interpreting and visualizing your SVEM models
If you would like to further explore Self-Validating Ensemble Models, check out the below resources: