[mlpack-git] master: Switch to random Acol initialization by default. (53a93a0)

[gitdub at big.cc.gt.atl.ga.us]

Repository : https://github.com/mlpack/mlpack

On branch  : master
Link       : https://github.com/mlpack/mlpack/compare/ca9c4c7d54e40516d87ccca3631a86148527b0a8...53a93a0e49ce0d472f1707b63f7659ceeb007c4f

>---------------------------------------------------------------

commit 53a93a0e49ce0d472f1707b63f7659ceeb007c4f
Author: Ryan Curtin <ryan at ratml.org>
Date:   Wed Jul 8 19:01:52 2015 +0000

    Switch to random Acol initialization by default.


>---------------------------------------------------------------

53a93a0e49ce0d472f1707b63f7659ceeb007c4f
 src/mlpack/methods/amf/amf.hpp | 32 +++++++++++++++++---------------
 1 file changed, 17 insertions(+), 15 deletions(-)

diff --git a/src/mlpack/methods/amf/amf.hpp b/src/mlpack/methods/amf/amf.hpp
index 1348ef5..1c90ea2 100644
--- a/src/mlpack/methods/amf/amf.hpp
+++ b/src/mlpack/methods/amf/amf.hpp
@@ -67,7 +67,7 @@ namespace amf /** Alternating Matrix Factorization **/ {
  * @see NMFMultiplicativeDistanceUpdate, SimpleResidueTermination
  */
 template<typename TerminationPolicyType = SimpleResidueTermination,
-         typename InitializationRuleType = RandomInitialization,
+         typename InitializationRuleType = RandomAcolInitialization,
          typename UpdateRuleType = NMFMultiplicativeDistanceUpdate>
 class AMF
 {
@@ -133,7 +133,7 @@ class AMF
 }; // class AMF
 
 typedef amf::AMF<amf::SimpleResidueTermination,
-                 amf::RandomInitialization,
+                 amf::RandomAcolInitialization,
                  amf::NMFALSUpdate> NMFALSFactorizer;
 
 //! Add simple typedefs
@@ -148,7 +148,7 @@ typedef amf::AMF<amf::SimpleResidueTermination,
  */
 template<class MatType>
 using SVDBatchFactorizer = amf::AMF<amf::SimpleResidueTermination,
-                                    amf::RandomInitialization,
+                                    amf::RandomAcolInitialization,
                                     amf::SVDBatchLearning>;
 
 /**
@@ -160,9 +160,10 @@ using SVDBatchFactorizer = amf::AMF<amf::SimpleResidueTermination,
  * @see SVDIncompleteIncrementalLearning
  */
 template<class MatType>
-using SVDIncompleteIncrementalFactorizer = amf::AMF<amf::SimpleResidueTermination,
-                                                    amf::RandomInitialization,
-                                                    amf::SVDIncompleteIncrementalLearning>;
+using SVDIncompleteIncrementalFactorizer = amf::AMF<
+    amf::SimpleResidueTermination,
+    amf::RandomAcolInitialization,
+    amf::SVDIncompleteIncrementalLearning>;
 /**
  * SVDCompleteIncrementalFactorizer factorizes given matrix V into two matrices
  * W and H by complete incremental gradient descent. SVD complete incremental
@@ -172,9 +173,10 @@ using SVDIncompleteIncrementalFactorizer = amf::AMF<amf::SimpleResidueTerminatio
  * @see SVDCompleteIncrementalLearning
  */
 template<class MatType>
-using SVDCompleteIncrementalFactorizer = amf::AMF<amf::SimpleResidueTermination,
-                                                  amf::RandomInitialization,
-                                                  amf::SVDCompleteIncrementalLearning<MatType> >;
+using SVDCompleteIncrementalFactorizer = amf::AMF<
+    amf::SimpleResidueTermination,
+    amf::RandomAcolInitialization,
+    amf::SVDCompleteIncrementalLearning<MatType>>;
 
 #else // #ifdef MLPACK_USE_CXX11
 
@@ -186,7 +188,7 @@ using SVDCompleteIncrementalFactorizer = amf::AMF<amf::SimpleResidueTermination,
  * @see SVDBatchLearning
  */
 typedef amf::AMF<amf::SimpleResidueTermination,
-                 amf::RandomInitialization,
+                 amf::RandomAcolInitialization,
                  amf::SVDBatchLearning> SparseSVDBatchFactorizer;
 
 /**
@@ -197,7 +199,7 @@ typedef amf::AMF<amf::SimpleResidueTermination,
  * @see SVDBatchLearning
  */
 typedef amf::AMF<amf::SimpleResidueTermination,
-                 amf::RandomInitialization,
+                 amf::RandomAcolInitialization,
                  amf::SVDBatchLearning> SVDBatchFactorizer;
 /**
  * SparseSVDIncompleteIncrementalFactorizer factorizes given sparse matrix V
@@ -208,7 +210,7 @@ typedef amf::AMF<amf::SimpleResidueTermination,
  * @see SVDIncompleteIncrementalLearning
  */
 typedef amf::AMF<amf::SimpleResidueTermination,
-                 amf::RandomInitialization,
+                 amf::RandomAcolInitialization,
                  amf::SVDIncompleteIncrementalLearning>
         SparseSVDIncompleteIncrementalFactorizer;
 
@@ -221,7 +223,7 @@ typedef amf::AMF<amf::SimpleResidueTermination,
  * @see SVDIncompleteIncrementalLearning
  */
 typedef amf::AMF<amf::SimpleResidueTermination,
-                 amf::RandomInitialization,
+                 amf::RandomAcolInitialization,
                  amf::SVDIncompleteIncrementalLearning>
         SVDIncompleteIncrementalFactorizer;
 
@@ -234,7 +236,7 @@ typedef amf::AMF<amf::SimpleResidueTermination,
  * @see SVDCompleteIncrementalLearning
  */
 typedef amf::AMF<amf::SimpleResidueTermination,
-                 amf::RandomInitialization,
+                 amf::RandomAcolInitialization,
                  amf::SVDCompleteIncrementalLearning<arma::sp_mat> >
         SparseSVDCompleteIncrementalFactorizer;
 
@@ -247,7 +249,7 @@ typedef amf::AMF<amf::SimpleResidueTermination,
  * @see SVDCompleteIncrementalLearning
  */
 typedef amf::AMF<amf::SimpleResidueTermination,
-                 amf::RandomInitialization,
+                 amf::RandomAcolInitialization,
                  amf::SVDCompleteIncrementalLearning<arma::mat> >
         SVDCompleteIncrementalFactorizer;