Problems with Laplace approximation method

[ismailsenoz]

There are couple of problems with the Laplace approximation:

1. For Laplace approximation we use Optim.jl package to evaluate derivatives, however in current version of Optim.jl derivative syntax for 1D functions has changed leading to failures in ReactiveMP.jl.
2. Current implementation of Laplace is not numerically stable.

[albertpod]

This isn't the first time we suffer from optimization packages. Should we use our handcrafted optimization procedures or we fix the version of optimization packages? Your thoughts guys, @ismailsenoz @bvdmitri @ismailsenoz?

[bvdmitri]

TBH I don't think we should use our handcrafted optimization procedures - maybe just another package? Or probably there is way to fix Optim.jl. There are plenty of other options too: GalacticOptim.

[albertpod]

Related #27