Add Apriori algorithm also for better comparison

[minakshikaushik]

No description provided.

[firefly-cpp]

Thanks, @minakshikaushik.

This is a good point. Although NiaARM is more or less oriented for numerical association rule mining, having a referential Apriori implementation in this repository would be beneficial for comparison, as you mentioned.

@zStupan, what do you think? Are you interested in adding basic Apriori implementation? Employing some classes from NiaARM may help speed up the performance. The discretization step exists in the tinynarm library.

By the way, there is a lack of pure Python Apriori implementations equipped with good documentation and still under active maintenance.

[firefly-cpp]

@zStupan, what do you think?

[zStupan]

Hello,

sorry for the late response, I don't really have a lot of free time right now... We could add an implementation of Apriori, although there are already some popular implementations out there (efficient-apriori).

So if I understand correctly, the comparison would be our method vs discretization and then running apriori?

[firefly-cpp]

@minakshikaushik, what do you think/need?

A comparison that may be suitable is our method using discrete dataset vs. apriori in terms of the number of rules generated, supports, confidences and time (very important), etc.

@zStupan, do you think including any existing implementations may be more convenient?

[minakshikaushik]

@firefly-cpp
Your idea "our method using discrete dataset vs. apriori in terms of the number of rules generated, supports, confidences and time (very important), etc." is good.

[firefly-cpp]

@zStupan, any progress, maybe?