AveragedLinearOptions.xml

<Type Name="AveragedLinearOptions" FullName="Microsoft.ML.Trainers.AveragedLinearOptions">
  <TypeSignature Language="C#" Value="public abstract class AveragedLinearOptions : Microsoft.ML.Trainers.OnlineLinearOptions" />
  <TypeSignature Language="ILAsm" Value=".class public auto ansi abstract beforefieldinit AveragedLinearOptions extends Microsoft.ML.Trainers.OnlineLinearOptions" />
  <TypeSignature Language="DocId" Value="T:Microsoft.ML.Trainers.AveragedLinearOptions" />
  <TypeSignature Language="VB.NET" Value="Public MustInherit Class AveragedLinearOptions&#xA;Inherits OnlineLinearOptions" />
  <TypeSignature Language="F#" Value="type AveragedLinearOptions = class&#xA;    inherit OnlineLinearOptions" />
  <AssemblyInfo>
    <AssemblyName>Microsoft.ML.StandardTrainers</AssemblyName>
    <AssemblyVersion>1.0.0.0</AssemblyVersion>
  </AssemblyInfo>
  <Base>
    <BaseTypeName>Microsoft.ML.Trainers.OnlineLinearOptions</BaseTypeName>
  </Base>
  <Interfaces />
  <Docs>
    <summary>
            Arguments class for averaged linear trainers.
            </summary>
    <remarks>To be added.</remarks>
  </Docs>
  <Members>
    <Member MemberName=".ctor">
      <MemberSignature Language="C#" Value="protected AveragedLinearOptions ();" />
      <MemberSignature Language="ILAsm" Value=".method familyhidebysig specialname rtspecialname instance void .ctor() cil managed" />
      <MemberSignature Language="DocId" Value="M:Microsoft.ML.Trainers.AveragedLinearOptions.#ctor" />
      <MemberSignature Language="VB.NET" Value="Protected Sub New ()" />
      <MemberType>Constructor</MemberType>
      <AssemblyInfo>
        <AssemblyName>Microsoft.ML.StandardTrainers</AssemblyName>
        <AssemblyVersion>1.0.0.0</AssemblyVersion>
      </AssemblyInfo>
      <Parameters />
      <Docs>
        <summary>To be added.</summary>
        <remarks>To be added.</remarks>
      </Docs>
    </Member>
    <Member MemberName="Averaged">
      <MemberSignature Language="C#" Value="public bool Averaged;" />
      <MemberSignature Language="ILAsm" Value=".field public bool Averaged" />
      <MemberSignature Language="DocId" Value="F:Microsoft.ML.Trainers.AveragedLinearOptions.Averaged" />
      <MemberSignature Language="VB.NET" Value="Public Averaged As Boolean " />
      <MemberSignature Language="F#" Value="val mutable Averaged : bool" Usage="Microsoft.ML.Trainers.AveragedLinearOptions.Averaged" />
      <MemberType>Field</MemberType>
      <AssemblyInfo>
        <AssemblyName>Microsoft.ML.StandardTrainers</AssemblyName>
        <AssemblyVersion>1.0.0.0</AssemblyVersion>
      </AssemblyInfo>
      <ReturnValue>
        <ReturnType>System.Boolean</ReturnType>
      </ReturnValue>
      <Docs>
        <summary>
            Determines whether to do averaging or not.
            </summary>
        <returns>
          <see langword="true" /> to do averaging; otherwise, <see langword="false" />.
            Default is <see langword="true" />.
            </returns>
        <remarks>To be added.</remarks>
      </Docs>
    </Member>
    <Member MemberName="DecreaseLearningRate">
      <MemberSignature Language="C#" Value="public bool DecreaseLearningRate;" />
      <MemberSignature Language="ILAsm" Value=".field public bool DecreaseLearningRate" />
      <MemberSignature Language="DocId" Value="F:Microsoft.ML.Trainers.AveragedLinearOptions.DecreaseLearningRate" />
      <MemberSignature Language="VB.NET" Value="Public DecreaseLearningRate As Boolean " />
      <MemberSignature Language="F#" Value="val mutable DecreaseLearningRate : bool" Usage="Microsoft.ML.Trainers.AveragedLinearOptions.DecreaseLearningRate" />
      <MemberType>Field</MemberType>
      <AssemblyInfo>
        <AssemblyName>Microsoft.ML.StandardTrainers</AssemblyName>
        <AssemblyVersion>1.0.0.0</AssemblyVersion>
      </AssemblyInfo>
      <Attributes>
        <Attribute FrameworkAlternate="ml-dotnet">
          <AttributeName Language="C#">[Microsoft.ML.SearchSpace.BooleanChoice]</AttributeName>
          <AttributeName Language="F#">[&lt;Microsoft.ML.SearchSpace.BooleanChoice&gt;]</AttributeName>
        </Attribute>
      </Attributes>
      <ReturnValue>
        <ReturnType>System.Boolean</ReturnType>
      </ReturnValue>
      <Docs>
        <summary>
            Determine whether to decrease the <see cref="F:Microsoft.ML.Trainers.AveragedLinearOptions.LearningRate" /> or not.
            </summary>
        <returns>
          <see langword="true" /> to decrease the <see cref="F:Microsoft.ML.Trainers.AveragedLinearOptions.LearningRate" /> as iterations progress; otherwise, <see langword="false" />.
            Default is <see langword="false" />. The learning rate will be reduced with every weight update proportional to the square root of the number of updates.
            </returns>
        <remarks>To be added.</remarks>
      </Docs>
    </Member>
    <Member MemberName="L2Regularization">
      <MemberSignature Language="C#" Value="public float L2Regularization;" />
      <MemberSignature Language="ILAsm" Value=".field public float32 L2Regularization" />
      <MemberSignature Language="DocId" Value="F:Microsoft.ML.Trainers.AveragedLinearOptions.L2Regularization" />
      <MemberSignature Language="VB.NET" Value="Public L2Regularization As Single " />
      <MemberSignature Language="F#" Value="val mutable L2Regularization : single" Usage="Microsoft.ML.Trainers.AveragedLinearOptions.L2Regularization" />
      <MemberType>Field</MemberType>
      <AssemblyInfo>
        <AssemblyName>Microsoft.ML.StandardTrainers</AssemblyName>
        <AssemblyVersion>1.0.0.0</AssemblyVersion>
      </AssemblyInfo>
      <Attributes>
        <Attribute FrameworkAlternate="ml-dotnet">
          <AttributeName Language="C#">[Microsoft.ML.SearchSpace.Range(0, 32768, 0, false)]</AttributeName>
          <AttributeName Language="F#">[&lt;Microsoft.ML.SearchSpace.Range(0, 32768, 0, false)&gt;]</AttributeName>
        </Attribute>
      </Attributes>
      <ReturnValue>
        <ReturnType>System.Single</ReturnType>
      </ReturnValue>
      <Docs>
        <summary>
            The L2 weight for <a href="https://en.wikipedia.org/wiki/Regularization_(mathematics)">regularization</a>.
            </summary>
        <remarks>To be added.</remarks>
      </Docs>
    </Member>
    <Member MemberName="LazyUpdate">
      <MemberSignature Language="C#" Value="public bool LazyUpdate;" />
      <MemberSignature Language="ILAsm" Value=".field public bool LazyUpdate" />
      <MemberSignature Language="DocId" Value="F:Microsoft.ML.Trainers.AveragedLinearOptions.LazyUpdate" />
      <MemberSignature Language="VB.NET" Value="Public LazyUpdate As Boolean " />
      <MemberSignature Language="F#" Value="val mutable LazyUpdate : bool" Usage="Microsoft.ML.Trainers.AveragedLinearOptions.LazyUpdate" />
      <MemberType>Field</MemberType>
      <AssemblyInfo>
        <AssemblyName>Microsoft.ML.StandardTrainers</AssemblyName>
        <AssemblyVersion>1.0.0.0</AssemblyVersion>
      </AssemblyInfo>
      <ReturnValue>
        <ReturnType>System.Boolean</ReturnType>
      </ReturnValue>
      <Docs>
        <summary>
            Determines when to update averaged weights.
            </summary>
        <returns>
          <see langword="true" /> to update averaged weights only when loss is nonzero.
            <see langword="false" /> to update averaged weights on every example.
            Default is <see langword="true" />.
            </returns>
        <remarks>To be added.</remarks>
      </Docs>
    </Member>
    <Member MemberName="LearningRate">
      <MemberSignature Language="C#" Value="public float LearningRate;" />
      <MemberSignature Language="ILAsm" Value=".field public float32 LearningRate" />
      <MemberSignature Language="DocId" Value="F:Microsoft.ML.Trainers.AveragedLinearOptions.LearningRate" />
      <MemberSignature Language="VB.NET" Value="Public LearningRate As Single " />
      <MemberSignature Language="F#" Value="val mutable LearningRate : single" Usage="Microsoft.ML.Trainers.AveragedLinearOptions.LearningRate" />
      <MemberType>Field</MemberType>
      <AssemblyInfo>
        <AssemblyName>Microsoft.ML.StandardTrainers</AssemblyName>
        <AssemblyVersion>1.0.0.0</AssemblyVersion>
      </AssemblyInfo>
      <Attributes>
        <Attribute FrameworkAlternate="ml-dotnet">
          <AttributeName Language="C#">[Microsoft.ML.SearchSpace.Range(0.0001, 1, 1, true)]</AttributeName>
          <AttributeName Language="F#">[&lt;Microsoft.ML.SearchSpace.Range(0.0001, 1, 1, true)&gt;]</AttributeName>
        </Attribute>
      </Attributes>
      <ReturnValue>
        <ReturnType>System.Single</ReturnType>
      </ReturnValue>
      <Docs>
        <summary>
            Learning rate.
            </summary>
        <remarks>To be added.</remarks>
      </Docs>
    </Member>
    <Member MemberName="RecencyGain">
      <MemberSignature Language="C#" Value="public float RecencyGain;" />
      <MemberSignature Language="ILAsm" Value=".field public float32 RecencyGain" />
      <MemberSignature Language="DocId" Value="F:Microsoft.ML.Trainers.AveragedLinearOptions.RecencyGain" />
      <MemberSignature Language="VB.NET" Value="Public RecencyGain As Single " />
      <MemberSignature Language="F#" Value="val mutable RecencyGain : single" Usage="Microsoft.ML.Trainers.AveragedLinearOptions.RecencyGain" />
      <MemberType>Field</MemberType>
      <AssemblyInfo>
        <AssemblyName>Microsoft.ML.StandardTrainers</AssemblyName>
        <AssemblyVersion>1.0.0.0</AssemblyVersion>
      </AssemblyInfo>
      <ReturnValue>
        <ReturnType>System.Single</ReturnType>
      </ReturnValue>
      <Docs>
        <summary>
            Extra weight given to more recent updates.
            </summary>
        <returns>
            Default is 0, i.e. no extra gain.
            </returns>
        <remarks>To be added.</remarks>
      </Docs>
    </Member>
    <Member MemberName="RecencyGainMultiplicative">
      <MemberSignature Language="C#" Value="public bool RecencyGainMultiplicative;" />
      <MemberSignature Language="ILAsm" Value=".field public bool RecencyGainMultiplicative" />
      <MemberSignature Language="DocId" Value="F:Microsoft.ML.Trainers.AveragedLinearOptions.RecencyGainMultiplicative" />
      <MemberSignature Language="VB.NET" Value="Public RecencyGainMultiplicative As Boolean " />
      <MemberSignature Language="F#" Value="val mutable RecencyGainMultiplicative : bool" Usage="Microsoft.ML.Trainers.AveragedLinearOptions.RecencyGainMultiplicative" />
      <MemberType>Field</MemberType>
      <AssemblyInfo>
        <AssemblyName>Microsoft.ML.StandardTrainers</AssemblyName>
        <AssemblyVersion>1.0.0.0</AssemblyVersion>
      </AssemblyInfo>
      <ReturnValue>
        <ReturnType>System.Boolean</ReturnType>
      </ReturnValue>
      <Docs>
        <summary>
            Determines whether <see cref="F:Microsoft.ML.Trainers.AveragedLinearOptions.RecencyGain" /> is multiplicative or additive.
            </summary>
        <returns>
          <see langword="true" /> means <see cref="F:Microsoft.ML.Trainers.AveragedLinearOptions.RecencyGain" /> is multiplicative.
            <see langword="false" /> means <see cref="F:Microsoft.ML.Trainers.AveragedLinearOptions.RecencyGain" /> is additive.
            Default is <see langword="false" />.
            </returns>
        <remarks>To be added.</remarks>
      </Docs>
    </Member>
    <Member MemberName="ResetWeightsAfterXExamples">
      <MemberSignature Language="C#" Value="public long? ResetWeightsAfterXExamples;" />
      <MemberSignature Language="ILAsm" Value=".field public valuetype System.Nullable`1&lt;int64&gt; ResetWeightsAfterXExamples" />
      <MemberSignature Language="DocId" Value="F:Microsoft.ML.Trainers.AveragedLinearOptions.ResetWeightsAfterXExamples" />
      <MemberSignature Language="VB.NET" Value="Public ResetWeightsAfterXExamples As Nullable(Of Long) " />
      <MemberSignature Language="F#" Value="val mutable ResetWeightsAfterXExamples : Nullable&lt;int64&gt;" Usage="Microsoft.ML.Trainers.AveragedLinearOptions.ResetWeightsAfterXExamples" />
      <MemberType>Field</MemberType>
      <AssemblyInfo>
        <AssemblyName>Microsoft.ML.StandardTrainers</AssemblyName>
        <AssemblyVersion>1.0.0.0</AssemblyVersion>
      </AssemblyInfo>
      <ReturnValue>
        <ReturnType>System.Nullable&lt;System.Int64&gt;</ReturnType>
      </ReturnValue>
      <Docs>
        <summary>
            Number of examples after which weights will be reset to the current average.
            </summary>
        <returns>
            Default is <see langword="null" />, which disables this feature.
            </returns>
        <remarks>To be added.</remarks>
      </Docs>
    </Member>
  </Members>
</Type>