grid search parametets

[charlesmartin14]

It is unclear what makes a good initial starting point for the SVM

We can grid search the parameters to find out

try testing different values of sigma, learning_rate, momentum, decay, N-iter, etc

see, for example, the variations in: EMF_RBM_Ominglot-20.ipynb EMF_RBM_Ominglot-200fast.ipynb EMF_RBM_Ominglot-20faster.ipynb

rbm = emf_rbm.EMF_RBM(verbose=True, monitor=True, learning_rate=0.1, n_iter=20,momentum=0.25, sigma=0.01, decay=0.1)

[charlesmartin14]

See how the parameters affect the convergence of the free energy

[basaks]

Set up the grid search in # 915fedf9dd. I need to run it now.