Created:
May 28, 2018 01:32 PM
| Last Modified: May 28, 2018 10:32 AM(5461 views)
Hi,
While using predicting screening to identify the contribution of each individual parameter among the data each time i run the program on the data different portion and contribution is observed.
Created:
May 28, 2018 09:59 PM
| Last Modified: May 28, 2018 6:59 PM(5447 views)
| Posted in reply to message from Alma1 05-28-2018
Hi,
That is correct. Becasue the predictor screening uses boostrap forests to generate the contributions, each time you run the procdure you may see different resutls. The larger the number of predictors, the more variblity you will see in the results. I often run the procedure 5 times and take the top contributors from each round and compare or feed them forward to the next step of my analysis.
For more info on the platform here is the overview help page: