search

jirifilip / pyARC

An implementation of CBA (Classification Based on Assocation) algorithm
23 stars 25 forks source link

can you share more about algorithms used : for example about M1Algorithm, M2Algorithm #7

Closed Sandy4321 closed 4 years ago

Sandy4321 commented 4 years ago

can you share more about algorithms used

for example about M1Algorithm, M2Algorithm how to find description about them in Liu, B. Hsu, W. and Ma, Y (1998). Integrating Classification and Association Rule Mining. Proceedings KDD-98, New York, 27-31 August. AAAI. pp 80-86.

KLIEGR, Tomas. Quantitative CBA: Small and Comprehensible Association Rule Classification Models. arXiv preprint arXiv:1711.10166, 2017.

per your code pred = m1clf.predict_all(txns_test)

predM2 = m2clf.predict_all(txns_test)

from https://github.com/jirifilip/pyARC/blob/master/notebooks/extensions/benchmarks/qcba_accuracy_benchmark.ipynb

kliegr commented 4 years ago

M1 and M2 are two algorithmic realizations of CBA, as described here: https://www.aaai.org/Papers/KDD/1998/KDD98-012.pdf Both algorithms provide the same results differ in speed. In this paper http://ceur-ws.org/Vol-2204/paper6.pdf, we found that M1 implementation from pyARC is faster than the M2 implementation.

Sandy4321 commented 4 years ago

I see Thanks for references Will try to read Hopefully will be able to understand ideas Since I only started to learn this great techniques By the way do you have some simple itroductury level description May be something like material for students?

kliegr commented 4 years ago

Maybe you can refer to this tutorial: From association rules to interpretable classification models - a tutorial