Does anyone have a good example of using SVEM that significantly improves model criteria (e.g., R-squared, AIC, etc.) compared to other methodologies(stepwise regression) ?
While keeping the model construct the same as RSM(main, quadratic, second interaction). Specifically, I’m not looking to add more complex interaction terms due to the simplicity of the model.
@HyeWonJung, The advantage of SVEM is it can assess more model terms with less runs not necessarily improve the model performance. The technique is most useful for analyzing supersaturated DOEs. Phil and I gave a talk here at a past discovery that has some data sets. Re-Thinking the Design and Analysis of Experiments? (2021-EU-30MP-776) - JMP User Community
