Regular breakfast meetup: Tom Roth on Unsupervised Learning

This is a past event

112 people went

1 Munn Street

1 Munn Street · Sydney

How to find us

We are in the event space on level 5 of the Sydney AI Hub at Tank Stream Labs. Entry is on Level 4, and you can take the lift up.

Location image of event venue

Details

PLEASE NOTE THE NEW LOCATION - we are holding this meetup at the new Sydney AI Hub in Barangaroo / Miller's Point. There is an awesome event space, a great view, and Infopeople are still sponsoring breakfast, yay!

------------------------------------------------------------------------------------------------------

This is our regular breakfast meetup for data scientists, analysts, and just people interested in data to discuss their experiences and learn from each other.

This week we're really excited to have the brilliant Tom Roth share some of his learnings working with the latest unsupervised learning tools. Don't miss out!

****
7:30 am - Breakfast and Networking
8:00 - Food for Thought
8:05 - Tom Roth - Your Stars Cluster in Evil Signs
8:30 - Q&A
8:45 - Close

****
Talk Abstract:
Your Stars Cluster in Evil Signs: Many businesses use clustering techniques on their data. But how are they explaining their results to stakeholders? What techniques are they using? What's the whole point of clustering? And what's this “evil signs” business?

In this talk we go through a clustering workflow, from data preparation to final output. We look at some of the latest tools available, like UMAP and HDBSCAN, and how they compare against the competition. Finally, I give my answers to these important questions.

****
Speaker Bio:
Tom fell into data science by accident, looking for an alternative to a PhD. He's glad he did. Tom is now a data scientist at Verge Labs, a small startup building AI applications for companies. He has worked in data science across Australia's most mega megacorporations, including Telstra, Caltex and Woolworths. Tom likes to build things, also writing about them, and he hosts a data science website at www.puzzlr.org. Tom also likes to read an unhealthy amount, take long walks on the beach, and listen to strange music.