Need Some Information About "Weight by Rule" & "Information Gain" - HELP PLS

shahinexir

hello

Ive been working on my thesis and i need to work with "weight by rule" & "Weight by Information Gain" Operators in rapidminer.

but the problem is i really have no idea what does These two operators do in order to give each attribute a weight. i mean i dont know how do they work and how they calculate the weights of each attribute.

 

ive read the documantaion given by rapidminer website but the description was too short and still i cant get how do these two operators work.

 

so im asking you for any kind of information about these operators in rapidminer and explanation that how do they work. so any information would be appreciated . thank you so much . best regards 

MartinLiebig

Hi,

 

you may wanna have a look at the implementation.

see: https://github.com/rapidminer/rapidminer-studio/blob/master/src/main/java/com/rapidminer/operator/features/weighting/InfoGainRatioWeighting.java

+

https://github.com/rapidminer/rapidminer-studio/tree/master/src/main/java/com/rapidminer/operator/learner/tree/criterions

 

Best,

Martin

shahinexir

hi, 
thank you so much for your reply
but i dont need the implementation.

the only thing that i need to know is that how for example "weight by rule" operator works. what does it actually do on a attribute and how does it calculate the weight of each attr.

i just need a very clear and brief explanation of it and thats it . 

do you think you can help me with that ? i really appreciate it . thank you.

MartinLiebig

Hi,

 

Weight by Rule should implement the OneR algorithm. This explains it: http://www.saedsayad.com/oner.htm

 

But i've not double checked it.

 

Cheers,

Martin