"Weight of evidence transformation"

OlaOla Member Posts: 5 Contributor II
edited May 23 in Help
I recently read some books about credit scoring and discovered the benefits of Weight of evidence transformation of continuous variables.

Can I perform this process in Rapid Miner and if so, how would I go about doing it?

Thanks,

Ola
Tagged:

Answers

  • OlaOla Member Posts: 5 Contributor II
    What I'm looking for is apparently also called pooling algorithms like MAPA, described here:
    http://www.crc.man.ed.ac.uk/conference/archive/2005/presentations/anderson-raymond.pdf

    The purpose is to find a grouping that attains a monotonous relationship with the target variable.

    There must be something in Rapid Miner that ressembles this, no?  ???

    In Enterprise Miner, there is functionality for Interactive Binning or, in the credit scoring, Interactive Grouping

    Ideas anyone?


Sign In or Register to comment.