<?xml version="1.0" encoding="utf-8" standalone="no"?>
<!DOCTYPE document PUBLIC "-//CNX//DTD CNXML 0.5 plus MathML//EN" "http://cnx.rice.edu/cnxml/0.5/DTD/cnxml_mathml.dtd">
<document xmlns="http://cnx.rice.edu/cnxml" xmlns:md="http://cnx.rice.edu/mdml/0.4" xmlns:m="http://www.w3.org/1998/Math/MathML" xmlns:bib="http://bibtexml.sf.net/" id="new5">
  <name>Existence of the Minimum Variance Unbiased Estimator (MVUB)</name>
  <metadata>
  <md:version>1.2</md:version>
  <md:created>2003/06/24 12:14:01 GMT-5</md:created>
  <md:revised>2003/07/10 10:36:59.432 GMT-5</md:revised>
  <md:authorlist>
    <md:author id="nowak">
      <md:firstname>Rob</md:firstname>
      <md:othername>"The Kid"</md:othername>
      <md:surname>Nowak</md:surname>
      <md:email>nowak@rice.edu</md:email>
    </md:author>
    <md:author id="cscott">
      <md:firstname>Clayton</md:firstname>
      
      <md:surname>Scott</md:surname>
      <md:email>cscott@rice.edu</md:email>
    </md:author>
  </md:authorlist>

  <md:maintainerlist>
    <md:maintainer id="nowak">
      <md:firstname>Rob</md:firstname>
      <md:othername>"The Kid"</md:othername>
      <md:surname>Nowak</md:surname>
      <md:email>nowak@rice.edu</md:email>
    </md:maintainer>
    <md:maintainer id="cscott">
      <md:firstname>Clayton</md:firstname>
      
      <md:surname>Scott</md:surname>
      <md:email>cscott@rice.edu</md:email>
    </md:maintainer>
    <md:maintainer id="lizzardg">
      <md:firstname>Elizabeth</md:firstname>
      
      <md:surname>Gregory</md:surname>
      <md:email>lizzardg@rice.edu</md:email>
    </md:maintainer>
    <md:maintainer id="jsilv">
      <md:firstname>Jeffrey</md:firstname>
      
      <md:surname>Silverman</md:surname>
      <md:email>jsilv@rice.edu</md:email>
    </md:maintainer>
  </md:maintainerlist>
  
  <md:keywordlist>
    <md:keyword>minimum variance</md:keyword>
    <md:keyword>unbiased estimator</md:keyword>
    <md:keyword>MVUB</md:keyword>
  </md:keywordlist>

  <md:abstract/>
</metadata>

  <content>
    <para id="delete_me">
      Does an MVUB estimator exist?  Suppose there exist three unbiased estimators: 
      <m:math>
	<m:apply>
	  <m:csymbol definitionURL="http://cnx.rice.edu/cd/cnxmath.ocd#estimate"/>
	  <m:ci><m:msub>
	      <m:mi>θ</m:mi>
	      <m:mn>1</m:mn>
	    </m:msub></m:ci>
	</m:apply>
      </m:math>, 
      <m:math>
	<m:apply>
	  <m:csymbol definitionURL="http://cnx.rice.edu/cd/cnxmath.ocd#estimate"/>
	  <m:ci><m:msub>
	      <m:mi>θ</m:mi>
	      <m:mn>2</m:mn>
	    </m:msub></m:ci>
	</m:apply>
      </m:math>, and 
      <m:math>
	<m:apply>
	  <m:csymbol definitionURL="http://cnx.rice.edu/cd/cnxmath.ocd#estimate"/>
	  <m:ci><m:msub>
	      <m:mi>θ</m:mi>
	      <m:mn>3</m:mn>
	    </m:msub></m:ci>
	</m:apply>
      </m:math>.  Two possibilities exist, as shown in <cnxn target="fig1"/> and <cnxn target="fig2"/>. 
      <figure id="fig1">
	<media type="image/png" src=""/>
      </figure>
      <figure id="fig2">
	<media type="image/png" src=""/>
	<caption>No MVUB estimator exists!</caption>
      </figure>
    </para>
    <section id="sect1">
      <name>Finding the MVUB Estimator</name>
      <para id="para1">
	There is no simple, general procedure for finding the MVUB
	estimator.
      </para>
      <para id="para2">
	In the next several lectures we will discuss several
	approaches:
	<list id="list1" type="enumerated">
	  <item>Determine the so-called <term>Cramer-Rao
	  lowerbound</term> (CRLB) and verify that the estimator
	  achieves it.</item> <item>Apply the Rao-Blackwell theorem
	  (we talked about this earlier in the course).</item>
	  <item>Further restrict the estimator to a class of
	  estimators (<foreign>e.g.</foreign>, linear or polynomial
	  functions of the data).</item>
	</list>
      </para>
    </section>
    <section id="sect2">
      <name>Outline of Remainder of Course</name>
      <list id="list2" type="enumerated">
	<item>The Cramer-Rao Lower Bound</item>
	<item>Linear Statistical Models</item>
	<item>Maximum Likelihood Estimation</item>
	<item>Bayesian Estimation</item>
	<item>Waveform Estimation - Wiener and Kalman Filters</item>
	<item>Adaptive Filtering</item>
      </list>
    </section>
  </content>
  
</document>
