<?xml version="1.0" encoding="UTF-8"?>
<rss xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:taxo="http://purl.org/rss/1.0/modules/taxonomy/" version="2.0">
  <channel>
    <title>topic GMM starting values in Statistical Procedures</title>
    <link>https://communities.sas.com/t5/Statistical-Procedures/GMM-starting-values/m-p/453429#M23723</link>
    <description>&lt;P&gt;Hi,&lt;/P&gt;&lt;P&gt;I'm used to R but a newbie for SAS.&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;When it comes to GMM Estimation,&lt;/P&gt;&lt;P&gt;How SAS set its starting values of parameters to minimize criterion function?&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;For example, in R, I need to set moment conditions and starting values for parameters to minimize criterion function, but I find that SAS do not require users to set starting values, as far as I know.&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;Since estimates are very sensitive to starting values, I wonder how SAS sets its starting values so that I can get the same estimates between R and SAS.&lt;/P&gt;</description>
    <pubDate>Thu, 12 Apr 2018 02:20:51 GMT</pubDate>
    <dc:creator>yongseokkim</dc:creator>
    <dc:date>2018-04-12T02:20:51Z</dc:date>
    <item>
      <title>GMM starting values</title>
      <link>https://communities.sas.com/t5/Statistical-Procedures/GMM-starting-values/m-p/453429#M23723</link>
      <description>&lt;P&gt;Hi,&lt;/P&gt;&lt;P&gt;I'm used to R but a newbie for SAS.&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;When it comes to GMM Estimation,&lt;/P&gt;&lt;P&gt;How SAS set its starting values of parameters to minimize criterion function?&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;For example, in R, I need to set moment conditions and starting values for parameters to minimize criterion function, but I find that SAS do not require users to set starting values, as far as I know.&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;Since estimates are very sensitive to starting values, I wonder how SAS sets its starting values so that I can get the same estimates between R and SAS.&lt;/P&gt;</description>
      <pubDate>Thu, 12 Apr 2018 02:20:51 GMT</pubDate>
      <guid>https://communities.sas.com/t5/Statistical-Procedures/GMM-starting-values/m-p/453429#M23723</guid>
      <dc:creator>yongseokkim</dc:creator>
      <dc:date>2018-04-12T02:20:51Z</dc:date>
    </item>
    <item>
      <title>Re: GMM starting values</title>
      <link>https://communities.sas.com/t5/Statistical-Procedures/GMM-starting-values/m-p/454523#M23724</link>
      <description>&lt;P&gt;Hi, welcome to SAS. I assume by GMM you mean "generalized mixed models"? Which procedure are you using?&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;&lt;A href="http://go.documentation.sas.com/?cdcId=pgmsascdc&amp;amp;cdcVersion=9.4_3.3&amp;amp;docsetId=statug&amp;amp;docsetTarget=titlepage.htm&amp;amp;locale=en" target="_self"&gt;The documentation of all the procedures is online&lt;/A&gt; and quite extensive. The&amp;nbsp;procedure that most people use for generalized linear mixed models is&amp;nbsp;&lt;A href="http://go.documentation.sas.com/?cdcId=pgmsascdc&amp;amp;cdcVersion=9.4_3.3&amp;amp;docsetId=statug&amp;amp;docsetTarget=statug_glimmix_toc.htm&amp;amp;locale=en" target="_self"&gt;the GLIMMIX procedure.&lt;/A&gt;&amp;nbsp;You can use &lt;A href="http://go.documentation.sas.com/?cdcId=pgmsascdc&amp;amp;cdcVersion=9.4_3.3&amp;amp;docsetId=statug&amp;amp;docsetTarget=statug_glimmix_syntax20.htm&amp;amp;locale=en" target="_self"&gt;the PARMS statement&lt;/A&gt; to set initial values (and bounds) for parameters.&lt;/P&gt;</description>
      <pubDate>Mon, 16 Apr 2018 18:07:47 GMT</pubDate>
      <guid>https://communities.sas.com/t5/Statistical-Procedures/GMM-starting-values/m-p/454523#M23724</guid>
      <dc:creator>Rick_SAS</dc:creator>
      <dc:date>2018-04-16T18:07:47Z</dc:date>
    </item>
  </channel>
</rss>

