PR feedback.

Author: codemzs

docs/api-reference/algo-details-sdca.md

@@ -0,0 +1,25 @@
+### Training Algorithm Details
+This trainer is based on the Stochastic Dual Coordinate Ascent (SDCA) method, a
+state-of-the-art optimization technique for convex objective functions. The
+algorithm can be scaled for use on large out-of-memory data sets due to a
+semi-asynchronized implementation that supports multi-threading.
+        
+Convergence is underwritten by periodically enforcing synchronization between
+primal and dual variables in a separate thread. Several choices of loss
+functions are also provided.
+          
+Note that SDCA is a stochastic and streaming optimization algorithm. The result
+depends on the order of training data because the stopping tolerance is not
+tight enough. In strongly-convex optimization, the optimal solution is unique
+and therefore everyone eventually reaches the same place. Even in
+non-strongly-convex cases, you will get equally-good solutions from run to run.
+For reproducible results, it is recommended that one sets 'Shuffle' to False and
+'NumThreads' to 1. Elastic net regularization can be specified by the 'L2Const'
+and 'L1Threshold' parameters. Note that the 'L2Const' has an effect on the rate
+of convergence. In general, the larger the 'L2Const', the faster SDCA converges.
+
+For more information, see:
+* [Scaling Up Stochastic Dual Coordinate
+  Ascent.](https://www.microsoft.com/en-us/research/wp-content/uploads/2016/06/main-3.pdf)
+* [Stochastic Dual Coordinate Ascent Methods for Regularized Loss
+  Minimization.](http://www.jmlr.org/papers/volume14/shalev-shwartz13a/shalev-shwartz13a.pdf)

src/Microsoft.ML.StandardTrainers/Standard/SdcaRegression.cs

@@ -39,22 +39,12 @@ namespace Microsoft.ML.Trainers
     /// | Is caching required? | No |
     /// | Required NuGet in addition to Microsoft.ML | None |
     ///
-    /// ### Training Algorithm Details
-    /// This trainer is based on the Stochastic Dual Coordinate Ascent (SDCA) method, a state-of-the-art optimization technique for convex objective functions.
-    /// The algorithm can be scaled for use on large out-of-memory data sets due to a semi-asynchronized implementation that supports multi-threading.
-    /// Convergence is underwritten by periodically enforcing synchronization between primal and dual updates in a separate thread.
-    /// Several choices of loss functions are also provided.The SDCA method combines several of the best properties and capabilities of logistic regression and SVM algorithms.
-    /// Note that SDCA is a stochastic and streaming optimization algorithm. The results depends on the order of the training data.
-    /// For reproducible results, it is recommended that one sets 'Shuffle' to False and 'NumThreads' to 1.
-    /// Elastic net regularization can be specified by the 'L2Const' and 'L1Threshold' parameters. Note that the 'L2Const' has an effect on the rate of convergence.
-    /// In general, the larger the 'L2Const', the faster SDCA converges.
-    /// For more information, see: [Scaling Up Stochastic Dual Coordinate Ascent](https://www.microsoft.com/en-us/research/wp-content/uploads/2016/06/main-3.pdf ) and
-    /// [Stochastic Dual Coordinate Ascent Methods for Regularized Loss Minimization](http://www.jmlr.org/papers/volume14/shalev-shwartz13a/shalev-shwartz13a.pdf).
+    /// [!include[io](~/../docs/samples/docs/api-reference/algo-details-sdca.md)]
     /// ]]>
     /// </format>
     /// </remarks>
-    /// <seealso cref="Microsoft.ML.StandardTrainersCatalog.Sdca(Microsoft.ML.RegressionCatalog.RegressionTrainers,System.String,System.String,System.String,Microsoft.ML.Trainers.ISupportSdcaRegressionLoss,System.Nullable{System.Single},System.Nullable{System.Single},System.Nullable{System.Int32})"/>
-    /// <seealso cref="Microsoft.ML.StandardTrainersCatalog.Sdca(Microsoft.ML.RegressionCatalog.RegressionTrainers,Microsoft.ML.Trainers.SdcaRegressionTrainer.Options)"/>
+    /// <seealso cref="StandardTrainersCatalog.Sdca(RegressionCatalog.RegressionTrainers, string, string, string, ISupportSdcaRegressionLoss, float?, float?, int?)"/>
+    /// <seealso cref="StandardTrainersCatalog.Sdca(RegressionCatalog.RegressionTrainers, SdcaRegressionTrainer.Options)"/>
     /// <seealso cref="Options"/>
     public sealed class SdcaRegressionTrainer : SdcaTrainerBase<SdcaRegressionTrainer.Options, RegressionPredictionTransformer<LinearRegressionModelParameters>, LinearRegressionModelParameters>
     {

src/Microsoft.ML.StandardTrainers/StandardTrainersCatalog.cs

@@ -129,7 +129,7 @@ public static SgdNonCalibratedTrainer SgdNonCalibrated(this BinaryClassification
         }
 
         /// <summary>
-        /// Creates a <see cref="SdcaRegressionTrainer"/>, which predicts a target using a linear regression model.
+        /// Create <see cref="SdcaRegressionTrainer"/>, which predicts a target using a linear regression model.
         /// </summary>
         /// <param name="catalog">The regression catalog trainer object.</param>
         /// <param name="labelColumnName">The name of the label column. The column data must be <see cref="System.Single"/></param>
@@ -160,7 +160,7 @@ public static SdcaRegressionTrainer Sdca(this RegressionCatalog.RegressionTraine
         }
 
         /// <summary>
-        /// Creates a <see cref="SdcaRegressionTrainer"/>, which predicts a target using a linear regression model trained with advanced options.
+        /// Creates <see cref="SdcaRegressionTrainer"/> with advanced options, which predicts a target using a linear regression model trained.
         /// </summary>
         /// <param name="catalog">The regression catalog trainer object.</param>
         /// <param name="options">Trainer options.</param>
@@ -181,7 +181,7 @@ public static SdcaRegressionTrainer Sdca(this RegressionCatalog.RegressionTraine
         }
 
         /// <summary>
-        /// Creates a <see cref="SdcaLogisticRegressionBinaryTrainer"/>, that predicts a target using a linear classification model.
+        /// Create <see cref="SdcaLogisticRegressionBinaryTrainer"/>, that predicts a target using a linear classification model.
         /// </summary>
         /// <param name="catalog">The binary classification catalog trainer object.</param>
         /// <param name="labelColumnName">The name of the label column. The column data must be <see cref="System.Single"/>.</param>
@@ -211,7 +211,7 @@ public static SdcaLogisticRegressionBinaryTrainer SdcaLogisticRegression(
         }
 
         /// <summary>
-        /// Creates a <see cref="SdcaLogisticRegressionBinaryTrainer"/>, which predicts a target using a linear classification model with advanced options.
+        /// Create <see cref="SdcaLogisticRegressionBinaryTrainer"/> using advanced options, which predicts a target using a linear classification model.
         /// </summary>
         /// <param name="catalog">The binary classification catalog trainer object.</param>
         /// <param name="options">Trainer options.</param>