Andreas Tiffeau-Mayer
Results
3
comments of
Andreas Tiffeau-Mayer
The learning rate for `minimizeSPSA` is decreased in a non-exponential fashion, as follows: `a/(0.01*niter+k+1)**alpha` There is currently no capability to use a user-defined learning rate schedule, although that would be...
A simple way is to turn the noisy vector into a noisy scalar by combining the various objectives in some weighted manner. There is currently no functionality to find Pareto...
Adding some functionality of this sort sounds like a great idea! Would be happy to accept a pull-request implementing such an interface. Unfortunately currently this is not implemented.