Pages: [1]
  Print  
Author Topic: Attribute weighting and cross validation  (Read 1518 times)
Username
Newbie
*
Posts: 39


« on: July 24, 2008, 10:05:00 AM »

Hi,

I want to select e.g the 80% best attributes according to the information gain. The first inner operator of the WrapperXValidation operator must return AttributeWeights, so how can I modify my AttributeWeights returned from InfoGainWeighting that the weights of the 80% worst attributes are set to zero?

Thanks
Logged
Ingo Mierswa
Administrator
Hero Member
*****
Posts: 1226



WWW
« Reply #1 on: July 29, 2008, 11:29:00 AM »

Hi,

since the info gain weighting would deliver always the same weights I do not see why to use the Wrapper validation for that. You could simply apply the operator AttributeWeightSelection for using the top k attributes. Here is an example:

Code:
<operator name="Root" class="Process" expanded="yes">
    <operator name="ExampleSetGenerator" class="ExampleSetGenerator">
        <parameter key="number_examples" value="500"/>
        <parameter key="number_of_attributes" value="100"/>
        <parameter key="target_function" value="sum classification"/>
    </operator>
    <operator name="InfoGainWeighting" class="InfoGainWeighting">
    </operator>
    <operator name="AttributeWeightSelection" class="AttributeWeightSelection">
        <parameter key="k" value="80"/>
        <parameter key="weight_relation" value="top k"/>
    </operator>
</operator>

Hope that gives you the basic idea.

Cheers,
Ingo
Logged

Did you try our new Marketplace? Upload or download new Extensions, add comments, and organize your operators. Have a look at  http://marketplace.rapid-i.com
Pages: [1]
  Print  
 
Jump to: