0
$\begingroup$

I want to know wether the interaction between the continous variable A, the continous variable B and the group (factor with 3 levels) is associated with a continous outcome Y in a longitudinal setting. I know that the interaction between Group and A is associated with B. I tested

lme4::lmer(Y ~ A * B * Group + Time + (1|IDs))

against a model without that interaction term

lme4::lmer(Y ~ A : B + A : Group + B : Group + A + B + Group + Time + (1|IDs))

but was now told to use SVMs and e.g. check the importance of this interaction term A : B : Group. However, I am not sure wether SVMs work in this longitudinal context including a three-way interaction and a random effect (appart from the fact that the sample size is low). If they do, how? I have tried GPBoost to include the random effect, but it doesn't seem to account for such interactions. Any other suggestions?

Thank you :)

Improve this question
New contributor
Friedebert is a new contributor to this site. Take care in asking for clarification, commenting, and answering. Check out our Code of Conduct.
$\endgroup$
1
  • $\begingroup$ Why do you think GPBoost does not not include this interaction? $\endgroup$
    – Michael M
    Commented 6 hours ago

0

Reset to default

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.