# What is the difference between emerging and discriminative patterns?

I'm working on a way to make efficient classification on sequential patterns mining algorithms. Looking in the state-of-the-art I found that there is two type of algorithm to address this problem : discriminative and emerging.

Can someone help me understand :

1- what is the difference between those two type of algorithm ? 2- what is the easier to adapt on sequential patterns mining ?

Thanks for any reply.

-

## 1 Answer

It is two names for the same thing. Both are methods that retrieve pattern highly frequent in one class while not frequent in other classes.

For more information on such patterns, you can have a look at Characterizing Discriminative Patterns.

If you are not looking for speed, you can use a standard sequential pattern mining algorithm and just compute a measure comparing the frequency of the computed sequential patterns in each class. The patterns having the highest value are then the emerging (or discriminative) patterns. This method allows to test several approaches and is very easy to implement.

If you are interested in performance, then the idea is to avoid exploring some part of the search space where you know that you will not be able to find any emerging pattern above a given threshold. Such algorithm may be much more efficient, but are highly dependent on your measure to discriminate pattern.

-