❤️ Keep posted on all our events by subscribing to our Meetup channel!

Clustering is one the fundamental tasks in data analysis, but getting to a satisfactory result is often hard and frustrating. It requires a good understanding of distance measures, clustering algorithms and their hyperparameters. New approaches to clustering try to make this easier by allowing the user to interact with clustering algorithms much more directly. They allow the user to express their interests through pairwise constraints, which are the result of answering simple pairwise questions (i.e. should these two instances be clustered together or not?). In this webinar, we'll talk about how algorithms can use this information to quickly obtain better clusterings. We will focus on one approach in particular, COBRAS, and demonstrate how it can be used to interactively cluster data. For more information: - LinkedIn: https://www.linkedin.com/company/data... - MeetUp: https://www.meetup.com/rootlabs-x/eve... - Website: https://dataroots.io/

Check out the video

Get in touch

For a coffee or chat ;)