search

corels / corels

Learning Certifiably Optimal Rule Lists
https://corels.eecs.harvard.edu/
GNU General Public License v3.0
172 stars 20 forks source link

did you compared your code with ICRM : Interpretable Classification Rule Mining Algorithm #11

Closed Sandy4321 closed 5 years ago

Sandy4321 commented 5 years ago

ICRM -- An Interpretable Classification Rule Mining Algorithm

https://www.researchgate.net/publication/232633984_A_Survey_of_Discretization_Techniques_Taxonomy_and_Empirical_Analysis_in_Supervised_Learning

code https://waikato.github.io/weka-wiki/packages/unofficial/

image

nlarusstone commented 5 years ago

We have not, as we are not a rule mining technique, but rather a rule list generation technique. Feel free to substitute ICRM for FPGrowth to use with CORELS though!

Sandy4321 commented 5 years ago

and what about this https://github.com/jirifilip/pyARC an implementation of CBA (Classification Based on Assocation) algorithm

nlarusstone commented 5 years ago

That's a greedy method, so we should equal or outperform it in all cases. Please let us know if you find that not to be the case!

Sandy4321 commented 4 years ago

Feel free to substitute ICRM for FPGrowth to use with CORELS though! can you help to understand what do you mean?