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.

Negative fitness in Heuristics Miner

Hi!

I am using ProM 6.7 (revision 35885) in order to load an event log in the form of this CSV file. I apply the "Convert CSV to XES" plugin with the default parameters (simply adding Case ID column as Case Column) and then I run the "Mine for a Heuristics Net using Heuristics Miner" with the default parameters. The resulting fitness of the discovered process model is -0.3936: a negative value...

You can see the overall process in this video.

I have no idea what is going on. Any idea?

Thank you!

Comments

  • Hi 'mvilar',

    Does the following discussion provide further insights?
    https://www.win.tue.nl/promforum/discussion/335/quality-dimentions

    I also know that a new version of the heuristics miner is developed, although I'm not sure about the status.

    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
  • mvilar
    edited October 2017
    Hi Buijs,

    Thank you for your answer!

    I followed the steps defined in this discussion, specifically:

    1. I have converted the heuristics net into a Petri net
    2. I have used the 'Replay a Log on Petri Net for Conformance Analysis'

    But I have got the following warning:
    The initial marking is an empty marking. If this is not intended, remove the currently existing InitialMarkingConnection object and then use "create Initial Marking" plugin to create a non-empty initial marking.

    You can see the overall process in this video.

    So, as I know what my initial marking is into the resulting Petri net (and the final marking, by the way), I have ran the 'Create Initial Marking' plugin over my Petri net and I have repeated the previous steps, obtaining a fitness of 0.8431 (as you can see in this video).

    I think all steps that I ran are correct, aren't they?

    So, can we conclude that there is an error calculating the fitness into the 'Mine for a Heuristics Net using Heuristics Miner'? What would be yout opinion?

    Thank you so much!
  • I agree, there is an error in the calculation and your steps seems logical.
    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
  • Thank you!
Sign In or Register to comment.