From b960134d71d744c56f555dfc2b5402fd9e92e0fa Mon Sep 17 00:00:00 2001
From: mhall <mhall@e0a1b77d-ad91-4216-81b1-defd5f83fa92>
Date: Wed, 25 Nov 2009 01:37:12 +0000
Subject: [PATCH] Forgot to update the javadoc.

git-svn-id: https://svn.scms.waikato.ac.nz/svn/weka/trunk/weka@6101 e0a1b77d-ad91-4216-81b1-defd5f83fa92
---
 src/main/java/weka/classifiers/functions/SPegasos.java | 10 +++++++++-
 1 file changed, 9 insertions(+), 1 deletion(-)

diff --git a/src/main/java/weka/classifiers/functions/SPegasos.java b/src/main/java/weka/classifiers/functions/SPegasos.java
index 0752294c..de5b6e94 100644
--- a/src/main/java/weka/classifiers/functions/SPegasos.java
+++ b/src/main/java/weka/classifiers/functions/SPegasos.java
@@ -49,7 +49,7 @@ import weka.filters.unsupervised.attribute.Normalize;
 
 /**
  <!-- globalinfo-start -->
- * Implements the stochastic variant of the Pegasos (Primal Estimated sub-GrAdient SOlver for SVM) method of Shalev-Shwartz et al. (2007). This implementation globally replaces all missing values and transforms nominal attributes into binary ones. It also normalizes all attributes, so the coefficients in the output are based on the normalized data. For more information, see<br/>
+ * Implements the stochastic variant of the Pegasos (Primal Estimated sub-GrAdient SOlver for SVM) method of Shalev-Shwartz et al. (2007). This implementation globally replaces all missing values and transforms nominal attributes into binary ones. It also normalizes all attributes, so the coefficients in the output are based on the normalized data. Can either minimize the hinge loss (SVM) or log loss (logistic regression). For more information, see<br/>
  * <br/>
  * S. Shalev-Shwartz, Y. Singer, N. Srebro: Pegasos: Primal Estimated sub-GrAdient SOlver for SVM. In: 24th International Conference on MachineLearning, 807-814, 2007.
  * <p/>
@@ -72,6 +72,10 @@ import weka.filters.unsupervised.attribute.Normalize;
  <!-- options-start -->
  * Valid options are: <p/>
  * 
+ * <pre> -F
+ *  Set the loss function to minimize. 0 = hinge loss (SVM), 1 = log loss (logistic regression).
+ *  (default = 0)</pre>
+ * 
  * <pre> -L &lt;double&gt;
  *  The lambda regularization constant (default = 0.0001)</pre>
  * 
@@ -340,6 +344,10 @@ public class SPegasos extends AbstractClassifier
    <!-- options-start -->
    * Valid options are: <p/>
    * 
+   * <pre> -F
+   *  Set the loss function to minimize. 0 = hinge loss (SVM), 1 = log loss (logistic regression).
+   *  (default = 0)</pre>
+   * 
    * <pre> -L &lt;double&gt;
    *  The lambda regularization constant (default = 0.0001)</pre>
    * 
-- 
2.11.4.GIT