To prevent spam users, you can only post on this forum after registration, which is by invitation. If you want to post on the forum, please send me a mail (h DOT m DOT w DOT verbeek AT tue DOT nl) and I'll send you an invitation in return for an account.
ETM in ProM Vs in repository
Hi !
I imported the ETM algorithm into Eclipse via the repository. I've executed it on and very simple event log and compared the resulting tree with the one I get with the ETM via ProM and I do not have the same results. Have the ETM plugin been improved in ProM ? (the resulting tree from ProM is better than the one from the eclipse implementation)
I imported the ETM algorithm into Eclipse via the repository. I've executed it on and very simple event log and compared the resulting tree with the one I get with the ETM via ProM and I do not have the same results. Have the ETM plugin been improved in ProM ? (the resulting tree from ProM is better than the one from the eclipse implementation)
Benjamin Dalmas
PhD Student - Blaise Pascal University - Clermont-Ferrand
PhD Student - Blaise Pascal University - Clermont-Ferrand
Comments
-
Dear Benjamin,
Thank you for testing the ETM
With 'ProM V' do you mean ProM 5.2? If so, then in ProM 5.2 there was no ETM so you are comparing it with another algorithm, probably the genetic miner.
If you mean ProM 6.5(.1) then the reason lies probably in the fact that the ETM is an evolutionary algorithm so no two runs of the ETM are the same. When run long enough the end result wil be the same, but differences do occur.
Another explanation could be a differences in the parameter settings used.
Or of course a difference in the implementation, which actually degraded performance.
I hope this helps!Joos Buijs
Senior Data Scientist and process mining expert at APG (Dutch pension fund executor).
Previously Assistant Professor in Process Mining at Eindhoven University of Technology -
Dear Joos,
I'm testing it on ProM 6.5. It's weird since I've not changed anything in the source code, just launched the ETM as it is in the repository.
Even with 5000 generations, I do not get results as good as the ones I get with only 1000 generation in ProM, not matter how many time I try. The other parameters are the same for both executions.
Maybe I should try to increase even more the number of generations. I'll keep you informed !
Thanks for your helpBenjamin Dalmas
PhD Student - Blaise Pascal University - Clermont-Ferrand
Howdy, Stranger!
Categories
- 1.6K All Categories
- 45 Announcements / News
- 225 Process Mining
- 6 - BPI Challenge 2020
- 9 - BPI Challenge 2019
- 24 - BPI Challenge 2018
- 27 - BPI Challenge 2017
- 8 - BPI Challenge 2016
- 68 Research
- 1K ProM 6
- 394 - Usage
- 288 - Development
- 9 RapidProM
- 1 - Usage
- 7 - Development
- 54 ProM5
- 19 - Usage
- 187 Event Logs
- 32 - ProMimport
- 75 - XESame