I'm working on modeling musical score patterns into a systematized data warehouse, first trying SQL to try to search for repeating patterns. Although SQL discovery does work, it's very noisy & the set-oriented code gets pretty obtuse.

So I'm looking at mining these sequential patterns using better algorithms, but am overwhelmed by all the options & the modern DM vernacular (just picked up a couple of beginners DM books).

Which approach would I anticipate using when working with potentially noisy sets?