<?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 Re: Model Selection in SAS Enterprise Guide and SAS Enterprise Miner in SAS Data Science</title>
    <link>https://communities.sas.com/t5/SAS-Data-Science/Model-Selection-in-SAS-Enterprise-Guide-and-SAS-Enterprise-Miner/m-p/463415#M7039</link>
    <description>&lt;P&gt;Thanks, Anna.&amp;nbsp; I'll check it out.&amp;nbsp; -- George Rezek&lt;/P&gt;</description>
    <pubDate>Fri, 18 May 2018 18:25:01 GMT</pubDate>
    <dc:creator>grezek</dc:creator>
    <dc:date>2018-05-18T18:25:01Z</dc:date>
    <item>
      <title>Model Selection in SAS Enterprise Guide and SAS Enterprise Miner</title>
      <link>https://communities.sas.com/t5/SAS-Data-Science/Model-Selection-in-SAS-Enterprise-Guide-and-SAS-Enterprise-Miner/m-p/461828#M7035</link>
      <description>&lt;P&gt;In Enterprise Miner I used Interactive grouping to binning my input variables.&amp;nbsp; I connected to the unbinned variables the regression node and ran a logistic regression.&amp;nbsp; Then from the Interactive Grouping node I ran a logistic regression and a scorecard.&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;The results were ROC for the logistic regression without binning&amp;nbsp;was .839;&amp;nbsp; an improvement was seen in the logistic regression run on the binned variables (ROC = .892);&amp;nbsp; the scorecard run off of the binned variables gave an ROC of .888.&amp;nbsp; (see attached image)&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;I expected the binned variables to give a higher ROC than unbinned, but I'm unsure why the regression and scorecard yielded different results.&amp;nbsp; The Model Comparison node compares the output statistics very conveniently;&amp;nbsp;is there a way to compare the relevant settings&amp;nbsp;between the regression and scorecard.&amp;nbsp;&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;Thanks in advance for any suggestions.&amp;nbsp; -- George Rezek&lt;/P&gt;</description>
      <pubDate>Sun, 13 May 2018 11:23:58 GMT</pubDate>
      <guid>https://communities.sas.com/t5/SAS-Data-Science/Model-Selection-in-SAS-Enterprise-Guide-and-SAS-Enterprise-Miner/m-p/461828#M7035</guid>
      <dc:creator>grezek</dc:creator>
      <dc:date>2018-05-13T11:23:58Z</dc:date>
    </item>
    <item>
      <title>Re: Model Selection in SAS Enterprise Guide and SAS Enterprise Miner</title>
      <link>https://communities.sas.com/t5/SAS-Data-Science/Model-Selection-in-SAS-Enterprise-Guide-and-SAS-Enterprise-Miner/m-p/463358#M7036</link>
      <description>&lt;P&gt;HI&amp;nbsp;&lt;a href="https://communities.sas.com/t5/user/viewprofilepage/user-id/17169"&gt;@grezek&lt;/a&gt;,&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;Thanks for your question! Perhaps this webinar and community article will help here? &lt;A href="https://communities.sas.com/t5/Ask-the-Expert/Model-Selection-in-SAS-Enterprise-Guide-and-SAS-Enterprise-Miner/ta-p/336345" target="_blank"&gt;Model Selection in SAS Enterprise Guide and SAS Enterprise Miner&lt;/A&gt;&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;Best,&lt;/P&gt;
&lt;P&gt;Anna&lt;/P&gt;</description>
      <pubDate>Fri, 18 May 2018 14:46:10 GMT</pubDate>
      <guid>https://communities.sas.com/t5/SAS-Data-Science/Model-Selection-in-SAS-Enterprise-Guide-and-SAS-Enterprise-Miner/m-p/463358#M7036</guid>
      <dc:creator>AnnaBrown</dc:creator>
      <dc:date>2018-05-18T14:46:10Z</dc:date>
    </item>
    <item>
      <title>Re: Model Selection in SAS Enterprise Guide and SAS Enterprise Miner</title>
      <link>https://communities.sas.com/t5/SAS-Data-Science/Model-Selection-in-SAS-Enterprise-Guide-and-SAS-Enterprise-Miner/m-p/463399#M7037</link>
      <description>&lt;P&gt;If you are just running the Regression and Scorecard nodes after the Interactive Grouping node with defaults, then one thing you might want to check are the inputs used in the Regression node - by default, it is using both the GRP_ and the WOE_ variables for each input whereas the Scorecard node is only using the WOE_ variables.&amp;nbsp; When I set Use=No for the GRP_ variables&amp;nbsp; in the variables editor for the Regression node, then my assessment results matched.&lt;/P&gt;</description>
      <pubDate>Fri, 18 May 2018 17:10:16 GMT</pubDate>
      <guid>https://communities.sas.com/t5/SAS-Data-Science/Model-Selection-in-SAS-Enterprise-Guide-and-SAS-Enterprise-Miner/m-p/463399#M7037</guid>
      <dc:creator>WendyCzika</dc:creator>
      <dc:date>2018-05-18T17:10:16Z</dc:date>
    </item>
    <item>
      <title>Re: Model Selection in SAS Enterprise Guide and SAS Enterprise Miner</title>
      <link>https://communities.sas.com/t5/SAS-Data-Science/Model-Selection-in-SAS-Enterprise-Guide-and-SAS-Enterprise-Miner/m-p/463401#M7038</link>
      <description>Thanks, Wendy! Now I get the same results are well. - George Rezek</description>
      <pubDate>Fri, 18 May 2018 17:21:48 GMT</pubDate>
      <guid>https://communities.sas.com/t5/SAS-Data-Science/Model-Selection-in-SAS-Enterprise-Guide-and-SAS-Enterprise-Miner/m-p/463401#M7038</guid>
      <dc:creator>grezek</dc:creator>
      <dc:date>2018-05-18T17:21:48Z</dc:date>
    </item>
    <item>
      <title>Re: Model Selection in SAS Enterprise Guide and SAS Enterprise Miner</title>
      <link>https://communities.sas.com/t5/SAS-Data-Science/Model-Selection-in-SAS-Enterprise-Guide-and-SAS-Enterprise-Miner/m-p/463415#M7039</link>
      <description>&lt;P&gt;Thanks, Anna.&amp;nbsp; I'll check it out.&amp;nbsp; -- George Rezek&lt;/P&gt;</description>
      <pubDate>Fri, 18 May 2018 18:25:01 GMT</pubDate>
      <guid>https://communities.sas.com/t5/SAS-Data-Science/Model-Selection-in-SAS-Enterprise-Guide-and-SAS-Enterprise-Miner/m-p/463415#M7039</guid>
      <dc:creator>grezek</dc:creator>
      <dc:date>2018-05-18T18:25:01Z</dc:date>
    </item>
  </channel>
</rss>

