πŸŽ‰ πŸŽ‰. RAPIDMINER 9.8 IS OUT!!! πŸŽ‰ πŸŽ‰

RapidMiner 9.8 continues to innovate in data science collaboration, connectivity and governance

CLICK HERE TO DOWNLOAD

Different results on diff. Computers with exactly same Setup and same Data

Fred12Fred12 Member Posts: 344   Unicorn
edited November 2018 in Help

hi,

I was using Bagging with Rotation FOrests with J48graft base learner... I used same parameter setups and same training/testing data.. however I run the process on differen PC's, on my home PC with better Processor but less RAM, and on another with small Ghz Processor but huge RAM...

I got 93.4 % on the second PC with more RAM, but on my home PC I got a result with 97.4%--- how can that be?

I noticed the size of the grafted J48 trees was much bigger, about the double the size... e.g about 1200 Trees, and on the other PC they had only about 600-800 on average size of tree...

Β 

why is that? I don't want to make my results dependable on the Computer that I use.. how is that even possible?Β 

Answers

  • Thomas_OttThomas_Ott RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 1,760   Unicorn

    Did you set them to the same seed?

  • Fred12Fred12 Member Posts: 344   Unicorn

    yes, I used same seed in Bagging (1992) and default Seed 1 for Rotation Forest, however, I could not set a seed for W-J48graft because there was no option for that.. maybe thats a reason? I wonder the results are still so different...

Sign In or Register to comment.