八度中fminunc的Java实现

Question

I am trying to find a java version of octave's fminunc (function minimization unconstrained) library in Java. The goal is to use it for logistic regression.

Currently, I am using a home-brewed version of gradiant descent for cost minimization and I would like to be able to use an already existing library to do that (in Java) for me. This is related to my effort of porting octave code that we have from the Coursera Machine Learning course to Java.

Answer 1

Ahh, here are a few things you can check for suitability:

http://dev.davidsoergel.com/trac/jlibsvm/ https://github.com/davejacobs/ml http://commons.apache.org/proper/commons-math/ http://liebke.github.io/incanter/optimize-api.html#incanter.optimize/non-linear-model

Here are a couple of threads on the same subject: https://groups.google.com/forum/#!topic/clojure/c7dxBWJn_X4 http://answers.yahoo.com/question/index?qid=20100227125757AA7faky

Answer 2

Thanks, Brill. I ended up using QNMinimizer from Stanford NLP Library.

Answer 3

If anyone found the previous two answers unsatisfactory, I have ported the Fmincg function to Java. The Fmincg function takes the same inputs as Fminunc, and it is has better performance for large datasets. https://github.com/Dan12/MachineLearning/blob/master/src/main/java/com/mycompany/maventest/Fmincg.java