What I m looking for is a function like this:
optimizeByLBFGS(CurrentParamters,GradientOfParamters,Loss/Cost) return: optimized paramters
Then I will use the returned optimized paramters and calculate the new loss/cost and feed them again ... and again... in this function. What can I use for this purpose?
Something like scipy.optimize.minimize in python for java?
Current paramters and gradients are availabe in a array / vector in the form of 100 rows and 1 column. Is something avaialbe
Thanks in advance!