Dear all, I have a question about the optimization method used for the mixed model in JMP/JMP pro
Normally mixed linear models do not come with a closed-form solution; therefore, they need a numerical method to be fitted – some of the known methods are the following:
- Newton-Raphson
- Broyden-Fletcher-Goldfarb-Shanno (better known as BFGS)
- LBFGS (which is a memory-efficient implementation of BFGS)
- Powell
- Basin hopping
- Newton conjugate gradient
Does anyone know the specific method that JMP and JMP pro uses to implement the linear mixed model fit?
