Matrix Profiles in Stumpy
Data Skeptic
00:00
How to Find Neighboring Subsequences
The first thing is for exploratory data analysis. In fact, to calculate these matrix profiles, you don't even technically need time. And so that's essentially what you need. With those two things, then you can compute the matrix profile for your time series. But the first three steps are, again, they sound quite simple, but they'reactually quite hard to get towards.
Transcript
Play full episode
