searchspace.html

来自「数据挖掘方面最新软件」· HTML 代码 · 共 785 行 · 第 1/3 页

<P>
<DD><DL>
</DL>
</DD>
</DL>
<HR>

<A NAME="resetOutlierStatus()"><!-- --></A><H3>
resetOutlierStatus</H3>
<PRE>
public void <B>resetOutlierStatus</B>()</PRE>
<DL>
<DD>This method resets the Outlier Status for all Objects in the Search room to have a clean start or to have a new identification of outliers with a separate method. As this zeros all boolean outlier statuses of all objects associated to this
 Searchroom and also zeros all outlier smooth factors, a current status list should be drawn down and stored somewhere before using this method.
 
 ATTN: As this only uses references to Objects associated to a Searchroom, in case more than one Searchroom uses a (fraction) range of objects, this might override the results from other detections for those objects. But it is encouraged to
 associate objects to only one SearchSpace and use duplications of objects with similar vektors in other SearchRooms.
<P>
<DD><DL>
</DL>
</DD>
</DL>
<HR>

<A NAME="radiusODSearch(double, double, com.rapidminer.operator.preprocessing.outlier.SearchObject, int)"><!-- --></A><H3>
radiusODSearch</H3>
<PRE>
public void <B>radiusODSearch</B>(double&nbsp;d,
                           double&nbsp;p,
                           <A HREF="../../../../../com/rapidminer/operator/preprocessing/outlier/SearchObject.html" title="class in com.rapidminer.operator.preprocessing.outlier">SearchObject</A>&nbsp;rObject,
                           int&nbsp;kindOfDistance)</PRE>
<DL>
<DD>BruteForce Radius Search to determine the outlier status of an object rObject of the type SearchObject this method takes d and p as parameters acc. to distance based DB(p,D)-Outlier (Knorr, Ng) and identifies an object as being an outlier, if
 more than a proportion p of the objects is more than distance D from rObject away.
 
 The simplest approach is to make a radius search for rObject and compare its distance to all other objects step by step with D (in this case d). If more than M = N(1-p) objects are within d, than rObject is not an Outlier, else it is. Although
 this is an approach with O(N^2) for all objects (it is O(N) for rObject), this prunes the search as soon as more than M objects are within d from rObject to get some improvement.
<P>
<DD><DL>
</DL>
</DD>
</DL>
<HR>

<A NAME="allRadiusSearch(double, double, int)"><!-- --></A><H3>
allRadiusSearch</H3>
<PRE>
public void <B>allRadiusSearch</B>(double&nbsp;d,
                            double&nbsp;p,
                            int&nbsp;kindOfDistance)</PRE>
<DL>
<DD>This method invokes the class method radiusODSearch on all objects in the SearchSpace (associated to this Searchroom via the listOfObjects vektor). radiusODSearch does a brute force distance Outlier test based on the parameters d and p for
 DB(p,d)-Outliers acc. to Knorr and Ng's approach to unify statistical Outlier tests. The result of the Outliertest is stored in the Objects themselves, e.g. each SearchObject knows its Outlier status (set recently, e.g. by this search) and can
 tell it by using the SearchObject's class method getOutlierStatus() (see there!)
 
 Added feature: prints progress on STDOUT for each 10% segment (app.) one hash "#" is printed to show progress if brute force should hit complexity boundaries (e.g. with a lot of dimensions as well as lots of objects). This also prints the
 parameters d and p and N for better understanding
<P>
<DD><DL>
</DL>
</DD>
</DL>
<HR>

<A NAME="getAverageDistanceMeasures(int)"><!-- --></A><H3>
getAverageDistanceMeasures</H3>
<PRE>
public double[] <B>getAverageDistanceMeasures</B>(int&nbsp;kindOfDistance)</PRE>
<DL>
<DD>Returns the average distances measures for the objects in the SearchSpace, calculating:
 
 <p>
 mean distance
 <p>
 standard deviation
 <p>
 variance
 
 The calculation is time consuming and should only be invoked if the data set is parsed for the first time (to get a feeling on it for statistical choices of parameters p and d for e.g. DB(p,d)-Outliers). It parses the objects matrix upper half
 to build an array of distances between objects (without doubling and without the distances of objects to themselves) which should be (n^2-n)/2 distances of value.
<P>
<DD><DL>

<DT><B>Returns:</B><DD>double[3] of mean, variance and standard deviation</DL>
</DD>
</DL>
<HR>

<A NAME="getAverageLOFMeasures()"><!-- --></A><H3>
getAverageLOFMeasures</H3>
<PRE>
public double[] <B>getAverageLOFMeasures</B>()</PRE>
<DL>
<DD>Returns the average LOF measures for the objects in the SearchSpace, calculating:
 
 <p>
 mean LOF
 <p>
 standard deviation
 <p>
 variance
<P>
<DD><DL>

<DT><B>Returns:</B><DD>double[3] of mean, variance and standard deviation</DL>
</DD>
</DL>
<HR>

<A NAME="getMaximumOutlierFactor()"><!-- --></A><H3>
getMaximumOutlierFactor</H3>
<PRE>
public double <B>getMaximumOutlierFactor</B>()</PRE>
<DL>
<DD>This method returns the maximum Outlier Factor of all SearchObjects in the 
 SearchSpace. Attn: Due to initializing, the outlier factors should be greater 
 or equal to zero.
<P>
<DD><DL>
</DL>
</DD>
</DL>
<HR>

<A NAME="findKdistanceContainers(com.rapidminer.operator.preprocessing.outlier.SearchObject, int)"><!-- --></A><H3>
findKdistanceContainers</H3>
<PRE>
public void <B>findKdistanceContainers</B>(<A HREF="../../../../../com/rapidminer/operator/preprocessing/outlier/SearchObject.html" title="class in com.rapidminer.operator.preprocessing.outlier">SearchObject</A>&nbsp;so,
                                    int&nbsp;kindOfDistance)</PRE>
<DL>
<DD><p>This method processes a sequential search over the SearchSpace for a SearchObject so 
 (named p here to be in line with the literature).</p>
 
 <p>As a result of the search a structure of k-distance-Containers is build and listed 
 within the SearchObject. Each container for a distance of an object or a number of 
 objects o in relation to p is filled with all the objects within that
 distance. The containers are sorted in a linked list in the SearchObject by 
 increasing distance. Just imagine it like p being a submarine sending a ping 
 and listing all echos in radiuses (=distance) with the echos stored in a band (=container)
 if they are on the same radius.</p>
<P>
<DD><DL>
<DT><B>Parameters:</B><DD><CODE>so</CODE> - </DL>
</DD>
</DL>
<HR>

<A NAME="findAllKdContainers(int)"><!-- --></A><H3>
findAllKdContainers</H3>
<PRE>
public void <B>findAllKdContainers</B>(int&nbsp;kindOfDistance)</PRE>
<DL>
<DD>Finds and fills all K distance containers for all objects in the Search Room by 
 invoking the process of finding all k distance containers for one Search Object.
<P>
<DD><DL>
</DL>
</DD>
</DL>
<HR>

<A NAME="computeLOF(int, int)"><!-- --></A><H3>
computeLOF</H3>
<PRE>
public void <B>computeLOF</B>(int&nbsp;kMin,
                       int&nbsp;kMax)</PRE>
<DL>
<DD><p>Some deeper magic to compute all the LOFs for the objects in the searchroom up to 
 MinPtsUB = kMax! The LOF output is only done up from kMin!</p>
 
 <p>This one is heavily documented in the source, so if you are interested on how it 
 is done, have a look at the source for the method.</p>
<P>
<DD><DL>
<DT><B>Parameters:</B><DD><CODE>kMin</CODE> - <DD><CODE>kMax</CODE> - </DL>
</DD>
</DL>
<HR>

<A NAME="computeDKN(int, int)"><!-- --></A><H3>
computeDKN</H3>
<PRE>
public void <B>computeDKN</B>(int&nbsp;dk,
                       int&nbsp;n)</PRE>
<DL>
<DD>This function computes the D^k_n Outliers according to Ramaswamy, Rastogi and 
 Shim which computes the top-n D^k-Outliers, the outliers (= objects) with the 
 maximum distance to the k-th nearest neighbors.
 
 Please be aware that this function requires the findAllKdContainers method has to be 
 run first, else it will simply stop or will not work.
<P>
<DD><DL>
<DT><B>Parameters:</B><DD><CODE>dk</CODE> - <DD><CODE>n</CODE> - </DL>
</DD>
</DL>
<!-- ========= END OF CLASS DATA ========= -->
<HR>


<!-- ======= START OF BOTTOM NAVBAR ====== -->
<A NAME="navbar_bottom"><!-- --></A>
<A HREF="#skip-navbar_bottom" title="Skip navigation links"></A>
<TABLE BORDER="0" WIDTH="100%" CELLPADDING="1" CELLSPACING="0" SUMMARY="">
<TR>
<TD COLSPAN=2 BGCOLOR="#EEEEFF" CLASS="NavBarCell1">
<A NAME="navbar_bottom_firstrow"><!-- --></A>
<TABLE BORDER="0" CELLPADDING="0" CELLSPACING="3" SUMMARY="">
  <TR ALIGN="center" VALIGN="top">
  <TD BGCOLOR="#EEEEFF" CLASS="NavBarCell1">    <A HREF="../../../../../overview-summary.html"><FONT CLASS="NavBarFont1"><B>Overview</B></FONT></A>&nbsp;</TD>
  <TD BGCOLOR="#EEEEFF" CLASS="NavBarCell1">    <A HREF="package-summary.html"><FONT CLASS="NavBarFont1"><B>Package</B></FONT></A>&nbsp;</TD>
  <TD BGCOLOR="#FFFFFF" CLASS="NavBarCell1Rev"> &nbsp;<FONT CLASS="NavBarFont1Rev"><B>Class</B></FONT>&nbsp;</TD>
  <TD BGCOLOR="#EEEEFF" CLASS="NavBarCell1">    <A HREF="class-use/SearchSpace.html"><FONT CLASS="NavBarFont1"><B>Use</B></FONT></A>&nbsp;</TD>
  <TD BGCOLOR="#EEEEFF" CLASS="NavBarCell1">    <A HREF="package-tree.html"><FONT CLASS="NavBarFont1"><B>Tree</B></FONT></A>&nbsp;</TD>
  <TD BGCOLOR="#EEEEFF" CLASS="NavBarCell1">    <A HREF="../../../../../deprecated-list.html"><FONT CLASS="NavBarFont1"><B>Deprecated</B></FONT></A>&nbsp;</TD>
  <TD BGCOLOR="#EEEEFF" CLASS="NavBarCell1">    <A HREF="../../../../../index-all.html"><FONT CLASS="NavBarFont1"><B>Index</B></FONT></A>&nbsp;</TD>
  <TD BGCOLOR="#EEEEFF" CLASS="NavBarCell1">    <A HREF="../../../../../help-doc.html"><FONT CLASS="NavBarFont1"><B>Help</B></FONT></A>&nbsp;</TD>
  </TR>
</TABLE>
</TD>
<TD ALIGN="right" VALIGN="top" ROWSPAN=3><EM>
<a href="http://sourceforge.net"><img src="http://sourceforge.net/sflogo.php?group_id=114160&amp;type=2" width="125" height="37" border="0" alt="SourceForge.net Logo" /></a></EM>
</TD>
</TR>

<TR>
<TD BGCOLOR="white" CLASS="NavBarCell2"><FONT SIZE="-2">
&nbsp;<A HREF="../../../../../com/rapidminer/operator/preprocessing/outlier/SearchObject.html" title="class in com.rapidminer.operator.preprocessing.outlier"><B>PREV CLASS</B></A>&nbsp;
&nbsp;NEXT CLASS</FONT></TD>
<TD BGCOLOR="white" CLASS="NavBarCell2"><FONT SIZE="-2">
  <A HREF="../../../../../index.html?com/rapidminer/operator/preprocessing/outlier/SearchSpace.html" target="_top"><B>FRAMES</B></A>  &nbsp;
&nbsp;<A HREF="SearchSpace.html" target="_top"><B>NO FRAMES</B></A>  &nbsp;
&nbsp;<SCRIPT type="text/javascript">
  <!--
  if(window==top) {
    document.writeln('<A HREF="../../../../../allclasses-noframe.html"><B>All Classes</B></A>');
  }
  //-->
</SCRIPT>
<NOSCRIPT>
  <A HREF="../../../../../allclasses-noframe.html"><B>All Classes</B></A>
</NOSCRIPT>


</FONT></TD>
</TR>
<TR>
<TD VALIGN="top" CLASS="NavBarCell3"><FONT SIZE="-2">
  SUMMARY:&nbsp;NESTED&nbsp;|&nbsp;FIELD&nbsp;|&nbsp;<A HREF="#constructor_summary">CONSTR</A>&nbsp;|&nbsp;<A HREF="#method_summary">METHOD</A></FONT></TD>
<TD VALIGN="top" CLASS="NavBarCell3"><FONT SIZE="-2">
DETAIL:&nbsp;FIELD&nbsp;|&nbsp;<A HREF="#constructor_detail">CONSTR</A>&nbsp;|&nbsp;<A HREF="#method_detail">METHOD</A></FONT></TD>
</TR>
</TABLE>
<A NAME="skip-navbar_bottom"></A>
<!-- ======== END OF BOTTOM NAVBAR ======= -->

<HR>
<i>Copyright &#169; 2001-2008 by <a href="http://rapid-i.com" target="_blank">Rapid-I</a></i>
</BODY>
</HTML>