thresholdselector.html

< 数据挖掘--实用机器学习技术及java实现> 一书结合数据挖掘和机器学习的知识,作者陈述了自动挖掘模式的基础理论,并且以java语言实现了具有代表性的各类数据挖掘方法.例如:class

<A NAME="TAGS_EVAL"><!-- --></A><H3>
TAGS_EVAL</H3>
<PRE>
public static final <A HREF="../../weka/core/Tag.html">Tag</A>[] <B>TAGS_EVAL</B></PRE>
<DL>
</DL>
<HR>

<A NAME="OPTIMIZE_0"><!-- --></A><H3>
OPTIMIZE_0</H3>
<PRE>
public static final int <B>OPTIMIZE_0</B></PRE>
<DL>
</DL>
<HR>

<A NAME="OPTIMIZE_1"><!-- --></A><H3>
OPTIMIZE_1</H3>
<PRE>
public static final int <B>OPTIMIZE_1</B></PRE>
<DL>
</DL>
<HR>

<A NAME="OPTIMIZE_LFREQ"><!-- --></A><H3>
OPTIMIZE_LFREQ</H3>
<PRE>
public static final int <B>OPTIMIZE_LFREQ</B></PRE>
<DL>
</DL>
<HR>

<A NAME="OPTIMIZE_MFREQ"><!-- --></A><H3>
OPTIMIZE_MFREQ</H3>
<PRE>
public static final int <B>OPTIMIZE_MFREQ</B></PRE>
<DL>
</DL>
<HR>

<A NAME="OPTIMIZE_POS_NAME"><!-- --></A><H3>
OPTIMIZE_POS_NAME</H3>
<PRE>
public static final int <B>OPTIMIZE_POS_NAME</B></PRE>
<DL>
</DL>
<HR>

<A NAME="TAGS_OPTIMIZE"><!-- --></A><H3>
TAGS_OPTIMIZE</H3>
<PRE>
public static final <A HREF="../../weka/core/Tag.html">Tag</A>[] <B>TAGS_OPTIMIZE</B></PRE>
<DL>
</DL>
<HR>

<A NAME="m_Classifier"><!-- --></A><H3>
m_Classifier</H3>
<PRE>
protected <A HREF="../../weka/classifiers/DistributionClassifier.html">DistributionClassifier</A> <B>m_Classifier</B></PRE>
<DL>
<DD>The generated base classifier</DL>
<HR>

<A NAME="m_HighThreshold"><!-- --></A><H3>
m_HighThreshold</H3>
<PRE>
protected double <B>m_HighThreshold</B></PRE>
<DL>
<DD>The upper threshold used as the basis of correction</DL>
<HR>

<A NAME="m_LowThreshold"><!-- --></A><H3>
m_LowThreshold</H3>
<PRE>
protected double <B>m_LowThreshold</B></PRE>
<DL>
<DD>The lower threshold used as the basis of correction</DL>
<HR>

<A NAME="m_BestThreshold"><!-- --></A><H3>
m_BestThreshold</H3>
<PRE>
protected double <B>m_BestThreshold</B></PRE>
<DL>
<DD>The threshold that lead to the best performance</DL>
<HR>

<A NAME="m_BestValue"><!-- --></A><H3>
m_BestValue</H3>
<PRE>
protected double <B>m_BestValue</B></PRE>
<DL>
<DD>The best value that has been observed</DL>
<HR>

<A NAME="m_NumXValFolds"><!-- --></A><H3>
m_NumXValFolds</H3>
<PRE>
protected int <B>m_NumXValFolds</B></PRE>
<DL>
<DD>The number of folds used in cross-validation</DL>
<HR>

<A NAME="m_Seed"><!-- --></A><H3>
m_Seed</H3>
<PRE>
protected int <B>m_Seed</B></PRE>
<DL>
<DD>Random number seed</DL>
<HR>

<A NAME="m_DesignatedClass"><!-- --></A><H3>
m_DesignatedClass</H3>
<PRE>
protected int <B>m_DesignatedClass</B></PRE>
<DL>
<DD>Designated class value, determined during building</DL>
<HR>

<A NAME="m_ClassMode"><!-- --></A><H3>
m_ClassMode</H3>
<PRE>
protected int <B>m_ClassMode</B></PRE>
<DL>
<DD>Method to determine which class to optimize for</DL>
<HR>

<A NAME="m_EvalMode"><!-- --></A><H3>
m_EvalMode</H3>
<PRE>
protected int <B>m_EvalMode</B></PRE>
<DL>
<DD>The evaluation mode</DL>
<HR>

<A NAME="m_RangeMode"><!-- --></A><H3>
m_RangeMode</H3>
<PRE>
protected int <B>m_RangeMode</B></PRE>
<DL>
<DD>The range correction mode</DL>
<HR>

<A NAME="MIN_VALUE"><!-- --></A><H3>
MIN_VALUE</H3>
<PRE>
protected static final double <B>MIN_VALUE</B></PRE>
<DL>
<DD>The minimum value for the criterion. If threshold adjustment
yields less than that, the default threshold of 0.5 is used.</DL>

<!-- ========= CONSTRUCTOR DETAIL ======== -->

<A NAME="constructor_detail"><!-- --></A>
<TABLE BORDER="1" CELLPADDING="3" CELLSPACING="0" WIDTH="100%">
<TR BGCOLOR="#CCCCFF" CLASS="TableHeadingColor">
<TD COLSPAN=1><FONT SIZE="+2">
<B>Constructor Detail</B></FONT></TD>
</TR>
</TABLE>

<A NAME="ThresholdSelector()"><!-- --></A><H3>
ThresholdSelector</H3>
<PRE>
public <B>ThresholdSelector</B>()</PRE>
<DL>
</DL>

<!-- ============ METHOD DETAIL ========== -->

<A NAME="method_detail"><!-- --></A>
<TABLE BORDER="1" CELLPADDING="3" CELLSPACING="0" WIDTH="100%">
<TR BGCOLOR="#CCCCFF" CLASS="TableHeadingColor">
<TD COLSPAN=1><FONT SIZE="+2">
<B>Method Detail</B></FONT></TD>
</TR>
</TABLE>

<A NAME="getPredictions(weka.core.Instances, int, int)"><!-- --></A><H3>
getPredictions</H3>
<PRE>
protected <A HREF="../../weka/core/FastVector.html">FastVector</A> <B>getPredictions</B>(<A HREF="../../weka/core/Instances.html">Instances</A>&nbsp;instances,
                                    int&nbsp;mode,
                                    int&nbsp;numFolds)
                             throws java.lang.Exception</PRE>
<DL>
<DD>Collects the classifier predictions using the specified evaluation method.<DD><DL>
</DL>
</DD>
<DD><DL>
<DT><B>Parameters:</B><DD><CODE>instances</CODE> - the set of <code>Instances</code> to generate
 predictions for.<DD><CODE>mode</CODE> - the evaluation mode.<DD><CODE>numFolds</CODE> - the number of folds to use if not evaluating on the
 full training set.<DT><B>Returns:</B><DD>a <code>FastVector</code> containing the predictions.<DT><B>Throws:</B><DD><CODE>java.lang.Exception</CODE> - if an error occurs generating the predictions.</DL>
</DD>
</DL>
<HR>

<A NAME="findThreshold(weka.core.FastVector)"><!-- --></A><H3>
findThreshold</H3>
<PRE>
protected void <B>findThreshold</B>(<A HREF="../../weka/core/FastVector.html">FastVector</A>&nbsp;predictions)</PRE>
<DL>
<DD>Finds the best threshold, this implementation searches for the
 highest FMeasure. If no FMeasure higher than MIN_VALUE is found,
 the default threshold of 0.5 is used.<DD><DL>
</DL>
</DD>
<DD><DL>
<DT><B>Parameters:</B><DD><CODE>predictions</CODE> - a <code>FastVector</code> containing the predictions.</DL>
</DD>
</DL>
<HR>

<A NAME="listOptions()"><!-- --></A><H3>
listOptions</H3>
<PRE>
public java.util.Enumeration <B>listOptions</B>()</PRE>
<DL>
<DD>Returns an enumeration describing the available options<DD><DL>
<DT><B>Specified by: </B><DD><CODE><A HREF="../../weka/core/OptionHandler.html#listOptions()">listOptions</A></CODE> in interface <CODE><A HREF="../../weka/core/OptionHandler.html">OptionHandler</A></CODE></DL>
</DD>
<DD><DL>
<DT><B>Returns:</B><DD>an enumeration of all the available options</DL>
</DD>
</DL>
<HR>

<A NAME="setOptions(java.lang.String[])"><!-- --></A><H3>
setOptions</H3>
<PRE>
public void <B>setOptions</B>(java.lang.String[]&nbsp;options)
                throws java.lang.Exception</PRE>
<DL>
<DD>Parses a given list of options. Valid options are:<p>

 -C num <br>
 The class for which threshold is determined. Valid values are:
 1, 2 (for first and second classes, respectively), 3 (for whichever
 class is least frequent), 4 (for whichever class value is most 
 frequent), and 5 (for the first class named any of "yes","pos(itive)",
 "1", or method 3 if no matches). (default 3). <p>

 -W classname <br>
 Specify the full class name of classifier to perform cross-validation
 selection on.<p>

 -X num <br> 
 Number of folds used for cross validation. If just a
 hold-out set is used, this determines the size of the hold-out set
 (default 3).<p>

 -R integer <br>
 Sets whether confidence range correction is applied. This can be used
 to ensure the confidences range from 0 to 1. Use 0 for no range correction,
 1 for correction based on the min/max values seen during threshold 
 selection (default 0).<p>

 -S seed <br>
 Random number seed (default 1).<p>

 -E integer <br>
 Sets the evaluation mode. Use 0 for evaluation using cross-validation,
 1 for evaluation using hold-out set, and 2 for evaluation on the
 training data (default 1).<p>

 Options after -- are passed to the designated sub-classifier. <p><DD><DL>
<DT><B>Specified by: </B><DD><CODE><A HREF="../../weka/core/OptionHandler.html#setOptions(java.lang.String[])">setOptions</A></CODE> in interface <CODE><A HREF="../../weka/core/OptionHandler.html">OptionHandler</A></CODE></DL>
</DD>
<DD><DL>
<DT><B>Parameters:</B><DD><CODE>options</CODE> - the list of options as an array of strings<DT><B>Throws:</B><DD><CODE>java.lang.Exception</CODE> - if an option is not supported</DL>
</DD>
</DL>
<HR>

<A NAME="getOptions()"><!-- --></A><H3>
getOptions</H3>
<PRE>
public java.lang.String[] <B>getOptions</B>()</PRE>
<DL>
<DD>Gets the current settings of the Classifier.<DD><DL>
<DT><B>Specified by: </B><DD><CODE><A HREF="../../weka/core/OptionHandler.html#getOptions()">getOptions</A></CODE> in interface <CODE><A HREF="../../weka/core/OptionHandler.html">OptionHandler</A></CODE></DL>
</DD>
<DD><DL>
<DT><B>Returns:</B><DD>an array of strings suitable for passing to setOptions</DL>
</DD>
</DL>
<HR>

<A NAME="buildClassifier(weka.core.Instances)"><!-- --></A><H3>
buildClassifier</H3>
<PRE>
public void <B>buildClassifier</B>(<A HREF="../../weka/core/Instances.html">Instances</A>&nbsp;instances)
                     throws java.lang.Exception</PRE>
<DL>
<DD>Generates the classifier.<DD><DL>
<DT><B>Overrides:</B><DD><CODE><A HREF="../../weka/classifiers/Classifier.html#buildClassifier(weka.core.Instances)">buildClassifier</A></CODE> in class <CODE><A HREF="../../weka/classifiers/Classifier.html">Classifier</A></CODE></DL>
</DD>
<DD><DL>
<DT><B>Parameters:</B><DD><CODE>instances</CODE> - set of instances serving as training data<DT><B>Throws:</B><DD><CODE>java.lang.Exception</CODE> - if the classifier has not been generated successfully</DL>
</DD>
</DL>
<HR>

<A NAME="distributionForInstance(weka.core.Instance)"><!-- --></A><H3>
distributionForInstance</H3>
<PRE>
public double[] <B>distributionForInstance</B>(<A HREF="../../weka/core/Instance.html">Instance</A>&nbsp;instance)
                                 throws java.lang.Exception</PRE>
<DL>
<DD>Calculates the class membership probabilities for the given test instance.<DD><DL>
<DT><B>Overrides:</B><DD><CODE><A HREF="../../weka/classifiers/DistributionClassifier.html#distributionForInstance(weka.core.Instance)">distributionForInstance</A></CODE> in class <CODE><A HREF="../../weka/classifiers/DistributionClassifier.html">DistributionClassifier</A></CODE></DL>
</DD>
<DD><DL>
<DT><B>Parameters:</B><DD><CODE>instance</CODE> - the instance to be classified<DT><B>Returns:</B><DD>predicted class probability distribution<DT><B>Throws:</B><DD><CODE>java.lang.Exception</CODE> - if instance could not be classified
 successfully</DL>
</DD>
</DL>
<HR>

<A NAME="globalInfo()"><!-- --></A><H3>
globalInfo</H3>
<PRE>
public java.lang.String <B>globalInfo</B>()</PRE>
<DL>
<DD><DL>
</DL>
</DD>
<DD><DL>
<DT><B>Returns:</B><DD>a description of the classifier suitable for
 displaying in the explorer/experimenter gui</DL>
</DD>
</DL>
<HR>

<A NAME="designatedClassTipText()"><!-- --></A><H3>
designatedClassTipText</H3>
<PRE>