machine learning - Is there a java libaray or code example availabe that optimize paramters via LBFGS by giving current parameters, gradient and cost/loss(function)? -
what m looking function this:
optimizebylbfgs(currentparamters,gradientofparamters,loss/cost) return: optimized paramters
then use returned optimized paramters , calculate new loss/cost , feed them again ... , again... in function. can use purpose?
something scipy.optimize.minimize in python java?
current paramters , gradients availabe in array / vector in form of 100 rows , 1 column. avaialbe
thanks in advance!
there 1 in spark: https://spark.apache.org/docs/1.0.1/api/java/org/apache/spark/mllib/optimization/lbfgs.html
and heard 1 nice too:
https://github.com/brendano/myutil/blob/master/src/util/lbfgs.java