› ›
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.

Question about the code of EvolutionaryTreeMiner

syjing628syjing628 Posts: 37
in - Development
Hey all, I am studying the code of ETM recently. I find that some guided mutation operators are diabled in the package "EvolutionaryTreeMiner", such as MutateOperatorTypeGuided, MutateLeafClassGuided, etc. I am confused that whether or not I should apply these operations in the ETM algorithm? Regards. Si-Yuan Jing
Tagged:

Comments

  • JBuijsJBuijs Posts: 914
    Hi Si-Yuan,

    Great to read that you're looking into the ETM and the guided mutators.

    It's  been a while for me (6 years) since I looked into the code so I'm a bit rusty. But if I recall correctly the code was either buggy/not working and/or had no positive effect in the ETM.

    You might find some more hints in my PhD thesis. And possibly the master thesis of Maikel van Eck.

    Please keep me updated if you're working on improving/extending the ETM, I see great potential there still.
    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
  • syjing628syjing628 Posts: 37
    Hi Joos, Thank you for your reply. ETM is a powerful tool and it shows great potential for process mining. There is another problem confused me. The algorithm outputs a very complex petri net when I convert a process tree to a petri net. I find that the class which is used to convert a process tree to a petri net is deprecated. Is there any bugs with the algorithm. Regards. Si-Yuan Jing
  • JBuijsJBuijs Posts: 914
    Hi Si-Yuan,

    It could be that in the past 6 years the process tree to petri net class became depricated and was replaced with a better/improved version, I do not know. I do recall however that the translations from process tree to Petri net indeed creates overly complex Petri nets, but apply reduction rules you should be able to reduce it properly.

    If I recall correctly Eric Verbeek and/or Sander Leemans might have plug-ins for this in ProM, maybe search for this?

    Good luck!
    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
  • hverbeekhverbeek Posts: 658
    Hi Si-Yuan,

    Yes, you can use the "Reduce Silent Transitions, Preserve Behavio" to reduce as much of the silent (black) transitions as possible. Better not to use the toehr reduction plugins, as they may change the behavior (the set of accepted traces).

    Kind regards,
    Eric.

  • syjing628syjing628 Posts: 37
    Hi Joos and Eric, Thanks for your response. It helps me a lot. Best regards, Si-Yuan Jing
Sign In or Register to comment.