Kernelized ridge regression -> Kernel ridge regression

Author: mblondel

doc/modules/classes.rst

@@ -580,7 +580,7 @@ From text
 
 .. _kernel_ridge_ref:
 
-:mod:`sklearn.kernel_ridge` Kernelized Ridge Regression
+:mod:`sklearn.kernel_ridge` Kernel Ridge Regression
 ========================================================
 
 .. automodule:: sklearn.kernel_ridge

doc/modules/kernel_ridge.rst

@@ -1,13 +1,13 @@
 .. _kernel_ridge:
 
 ===========================
-Kernelized ridge regression
+Kernel ridge regression
 ===========================
 
 .. currentmodule:: sklearn.kernel_ridge
 
-Kernelized ridge regression (KRR) [M2012]_ combines :ref:`ridge_regression` 
-(linear least squares plus l2-norm regularization) with the kernel trick. It 
+Kernel ridge regression (KRR) [M2012]_ combines :ref:`ridge_regression`
+(linear least squares with l2-norm regularization) with the kernel trick. It
 thus learns a linear function in the space induced by the respective kernel and
 the data. For non-linear kernels, this corresponds to a non-linear
 function in the original space.
@@ -16,9 +16,9 @@ The form of the model learned by :class:`KernelRidge` is identical to support
 vector regression (:class:`SVR`). However, different loss functions are used:
 KRR uses squared error loss while support vector regression uses
 :math:`\epsilon`-insensitive loss, both combined with l2 regularization.  In
-contrast to :class:`SVR`, fitting :class:`KernelRidge` can be done in 
-closed-form and is typically faster for medium-sized datasets. On the other 
-hand, the learned model is non-sparse and thus slower than SVR, which learns 
+contrast to :class:`SVR`, fitting :class:`KernelRidge` can be done in
+closed-form and is typically faster for medium-sized datasets. On the other
+hand, the learned model is non-sparse and thus slower than SVR, which learns
 a sparse model for :math:`\epsilon > 0`, at prediction-time.
 
 The following figure compares :class:`KernelRidge` and :class:`SVR` on

doc/supervised_learning.rst

@@ -8,6 +8,8 @@ Supervised learning
 .. toctree::
 
     modules/linear_model
+    modules/lda_qda.rst
+    modules/kernel_ridge.rst
     modules/svm
     modules/sgd
     modules/neighbors
@@ -19,6 +21,4 @@ Supervised learning
     modules/multiclass
     modules/feature_selection.rst
     modules/label_propagation.rst
-    modules/lda_qda.rst
     modules/isotonic.rst
-    modules/kernel_ridge.rst

sklearn/kernel_ridge.py

@@ -14,10 +14,10 @@
 
 
 class KernelRidge(BaseEstimator, RegressorMixin):
-    """Kernelized ridge regression.
+    """Kernel ridge regression.
 
-    Kernelized ridge regression (KRR) combines ridge regression (linear least
-    squares plus l2-norm regularization) with the kernel trick. It thus
+    Kernel ridge regression (KRR) combines ridge regression (linear least
+    squares with l2-norm regularization) with the kernel trick. It thus
     learns a linear function in the space induced by the respective kernel and
     the data. For non-linear kernels, this corresponds to a non-linear
     function in the original space.
@@ -82,7 +82,7 @@ class KernelRidge(BaseEstimator, RegressorMixin):
     See also
     --------
     Ridge
-        Linear, non-kernelized least squares with l2 regularization.
+        Linear ridge regression.
     SVR
         Support Vector Regression implemented using libsvm.