"Problems with the linear Regression"

Matthias

Hello,

this post refers to the Bug 497 I've send Friday. There I wrote that there ist a problem with the X- and split validation. Now a little upgrate: The validation operators are not the wrongdoer. The error occurs as well when I only use the linear regression operator and choose less than two attributes with the select attribute operator. When I generate two regular attributes with the generate data operator everything works fine?

<宏/ >

< =“tru运营商激活e" class="process" compatibility="5.1.002" expanded="true" name="Process">

< =“tru运营商激活e" class="generate_data" compatibility="5.1.002" expanded="true" height="60" name="Generate Data" width="90" x="45" y="210"/>
< =“tru运营商激活e" class="select_attributes" compatibility="5.1.002" expanded="true" height="76" name="Select Attributes (2)" width="90" x="179" y="300">



< =“tru运营商激活e" class="linear_regression" compatibility="5.1.002" expanded="true" height="94" name="Linear Regression (2)" width="90" x="313" y="255"/>

Because of this problem I can't use the forward selection operator too.

Best wishes

Matthias

land

Hi all,
this was a small bug in the LinearRegression that could happen if no feature was de-selected when using the internal feature selection. Another problem was the XValidation that by default threw an error if stratified sampling was chosen without a nominal label attribute present.

The newly released version fixes both issues. (Again someone was quite happy to be an enterprise customerSometimes this make things go faster...)


Greetings,
Sebastian