<?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: Activation Function on the HP Neural at SAS Miner in SAS Data Science</title>
    <link>https://communities.sas.com/t5/SAS-Data-Science/Activation-Function-on-the-HP-Neural-at-SAS-Miner/m-p/512710#M7500</link>
    <description>&lt;DIV class="lia-quilt-column lia-quilt-column-04 lia-quilt-column-left lia-quilt-column-main-left"&gt;&amp;nbsp;&lt;/DIV&gt;
&lt;DIV class="lia-quilt-column lia-quilt-column-20 lia-quilt-column-right lia-quilt-column-main-right"&gt;
&lt;DIV class="lia-quilt-column-alley lia-quilt-column-alley-right"&gt;
&lt;DIV id="messagebodydisplay_0" class="lia-message-body"&gt;
&lt;DIV class="lia-message-body-content"&gt;
&lt;BLOCKQUOTE&gt;
&lt;P&gt;Hi guys, how can I know what is the best activation function on the HP Neural Node at SAS Enterprise Miner?&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;In this node, I have the following activation functions: exponential, indetity, sine and tanh.&lt;/P&gt;
&lt;/BLOCKQUOTE&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;In Data Mining problems, you are relying on numerical methods to come up with answers that are not easily seen from the raw data due to the number of observations and/or variables.&amp;nbsp; The bad news is that there is no easy way to look at a new data set and decide what the best transformation is going to be.&amp;nbsp; &amp;nbsp;The good news is that it is easy to build multiple paths using different transformations for each so that you can quickly run and compare the approaches for the data in question.&amp;nbsp; This same type of approach addresses other related questions like "which tree settings should I use to get the best tree?" or "which predictive modeling method will work best for my data?".&amp;nbsp; &amp;nbsp;In the end, SAS Enterprise Miner allows you to easily build multiple competing flows which you can then compare so it is rarely if ever important to answer the question about which way is best before you begin.&amp;nbsp; If you continue to use various methods for a certain class of problems in your industry, you might find that certain approaches tend to work better than others for your business problem, but that does not mean they would be best for anyone else's business problem.&amp;nbsp;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;Hope this helps!&lt;/P&gt;
&lt;P&gt;Doug&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;/DIV&gt;
&lt;/DIV&gt;
&lt;/DIV&gt;
&lt;/DIV&gt;</description>
    <pubDate>Tue, 13 Nov 2018 19:34:33 GMT</pubDate>
    <dc:creator>DougWielenga</dc:creator>
    <dc:date>2018-11-13T19:34:33Z</dc:date>
    <item>
      <title>Activation Function on the HP Neural at SAS Miner</title>
      <link>https://communities.sas.com/t5/SAS-Data-Science/Activation-Function-on-the-HP-Neural-at-SAS-Miner/m-p/491348#M7328</link>
      <description>&lt;P&gt;Hi guys, how can I know what is the best activation function on the HP Neural Node at SAS Enterprise Miner?&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;In this node, I have the following activation functions: exponential, indetity, sine and tanh.&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;Thank's&lt;/P&gt;&lt;P&gt;&amp;nbsp;&lt;/P&gt;&lt;P&gt;Bryan&lt;/P&gt;</description>
      <pubDate>Thu, 30 Aug 2018 17:14:17 GMT</pubDate>
      <guid>https://communities.sas.com/t5/SAS-Data-Science/Activation-Function-on-the-HP-Neural-at-SAS-Miner/m-p/491348#M7328</guid>
      <dc:creator>bryanssilveira</dc:creator>
      <dc:date>2018-08-30T17:14:17Z</dc:date>
    </item>
    <item>
      <title>Re: Activation Function on the HP Neural at SAS Miner</title>
      <link>https://communities.sas.com/t5/SAS-Data-Science/Activation-Function-on-the-HP-Neural-at-SAS-Miner/m-p/512710#M7500</link>
      <description>&lt;DIV class="lia-quilt-column lia-quilt-column-04 lia-quilt-column-left lia-quilt-column-main-left"&gt;&amp;nbsp;&lt;/DIV&gt;
&lt;DIV class="lia-quilt-column lia-quilt-column-20 lia-quilt-column-right lia-quilt-column-main-right"&gt;
&lt;DIV class="lia-quilt-column-alley lia-quilt-column-alley-right"&gt;
&lt;DIV id="messagebodydisplay_0" class="lia-message-body"&gt;
&lt;DIV class="lia-message-body-content"&gt;
&lt;BLOCKQUOTE&gt;
&lt;P&gt;Hi guys, how can I know what is the best activation function on the HP Neural Node at SAS Enterprise Miner?&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;In this node, I have the following activation functions: exponential, indetity, sine and tanh.&lt;/P&gt;
&lt;/BLOCKQUOTE&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;In Data Mining problems, you are relying on numerical methods to come up with answers that are not easily seen from the raw data due to the number of observations and/or variables.&amp;nbsp; The bad news is that there is no easy way to look at a new data set and decide what the best transformation is going to be.&amp;nbsp; &amp;nbsp;The good news is that it is easy to build multiple paths using different transformations for each so that you can quickly run and compare the approaches for the data in question.&amp;nbsp; This same type of approach addresses other related questions like "which tree settings should I use to get the best tree?" or "which predictive modeling method will work best for my data?".&amp;nbsp; &amp;nbsp;In the end, SAS Enterprise Miner allows you to easily build multiple competing flows which you can then compare so it is rarely if ever important to answer the question about which way is best before you begin.&amp;nbsp; If you continue to use various methods for a certain class of problems in your industry, you might find that certain approaches tend to work better than others for your business problem, but that does not mean they would be best for anyone else's business problem.&amp;nbsp;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;Hope this helps!&lt;/P&gt;
&lt;P&gt;Doug&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;/DIV&gt;
&lt;/DIV&gt;
&lt;/DIV&gt;
&lt;/DIV&gt;</description>
      <pubDate>Tue, 13 Nov 2018 19:34:33 GMT</pubDate>
      <guid>https://communities.sas.com/t5/SAS-Data-Science/Activation-Function-on-the-HP-Neural-at-SAS-Miner/m-p/512710#M7500</guid>
      <dc:creator>DougWielenga</dc:creator>
      <dc:date>2018-11-13T19:34:33Z</dc:date>
    </item>
  </channel>
</rss>

