"{solved} Problem in visualization of weighting"
Hi .. i know we can get a Graph for weighting but my problem is i have so many attributes and i just need attributes which weights are more than .5
so for doing this i used select by weight then used data to weight operators ....but THE big problem is when i use it ....data to weight,weights all remaining attributes 1 ! which is not correct ..what should i do?
Second question: which operators are good or can be used for visualization...i already know that decision trees provide an excellent visual model....my job needs good graphs and model which are not so hard to explain yet they are attractive too
Danke schon
and here is xml code for my first Q
so for doing this i used select by weight then used data to weight operators ....but THE big problem is when i use it ....data to weight,weights all remaining attributes 1 ! which is not correct ..what should i do?
Second question: which operators are good or can be used for visualization...i already know that decision trees provide an excellent visual model....my job needs good graphs and model which are not so hard to explain yet they are attractive too
Danke schon
and here is xml code for my first Q
<宏/ >
<连接摇来摇去m_op="Retrieve" from_port="output" to_op="Select Attributes" to_port="example set input"/>
<连接from_op = "选择属性”from_port = "前女友ample set output" to_op="Multiply" to_port="input"/>
<连接摇来摇去m_op="Multiply" from_port="output 1" to_op="Weight by SAM" to_port="example set"/>
<连接摇来摇去m_op="Multiply" from_port="output 2" to_op="MR-Weighting" to_port="example set"/>
<连接摇来摇去m_op="MR-Weighting" from_port="weights" to_op="Select by Weights (2)" to_port="weights"/>
<连接摇来摇去m_op="MR-Weighting" from_port="example set" to_op="Select by Weights (2)" to_port="example set input"/>
<连接摇来摇去m_op="Weight by SAM" from_port="weights" to_op="Select by Weights" to_port="weights"/>
<连接摇来摇去m_op="Weight by SAM" from_port="example set" to_op="Select by Weights" to_port="example set input"/>
<连接摇来摇去m_op="Select by Weights" from_port="example set output" to_op="Data to Weights" to_port="example set"/>
<连接摇来摇去m_op="Data to Weights" from_port="weights" to_port="result 1"/>
<连接摇来摇去m_op="Select by Weights (2)" from_port="example set output" to_op="Data to Weights (2)" to_port="example set"/>
<连接摇来摇去m_op="Data to Weights (2)" from_port="weights" to_port="result 2"/>
Tagged:
0
Answers
What you need here is Weights to Data, OR a second weighting operator after removing the unimportant attributes. Please see the attached process for an example of the first alternative. It outputs the filtered example set, and the weights of the attributes in the original data set.
Second question: please open one post per question next time, unless the questions are *closely* related. Otherwise the threads will become cluttered, and the forum search won't deliver valuable results for the next user.
However: with "operators" here you probably mean learning algorithms: I have no idea, what you are doing with your data, but since you are mentioning decision trees, I guess you are dealing with classification tasks. In this case, e.g. the linear SVM delivers well understandable models with weights for each feature, similar to the Linear Regression, which can also be used for binary classification tasks. Naive Bayes of course also creates fast and easy models which can be nicely visualized.
Best, Marius
Best, Marius