In this episode of Science Talks, Snorkel AI’s Braden Hancock chats with Paroma Varma  -  a fellow co-founder at Snorkel AI, and one of the principal early contributors to the Snorkel project. We discuss Paroma’s path into machine learning, research into applying weak supervision to non-text domains, and the benefits of Snorkel Flow as an end-to-end AI pipeline.

ScienceTalks is an interview series from Snorkel AI, highlighting some of the best work and ideas to make AI practical.

More related videos: https://www.youtube.com/playlist?list=PLZePYakcDhmhU55eQQJux9y3DZyUxC5YC

00:00 – ScienceTalks Titles
00:16 – Introduction
00:48 – Paroma’s Path to ML
03:24 – Deciding on a PhD
04:58 – The PhD experience
07:16 – Weak Supervision in non-text domains
11:37 – Project Snuba
14:31 – The benefits of an end-to-end AI pipeline
19:16 – Working at Snorkel AI
20:52 – How to find Paroma

#machinelearning #weaksupervision #snorkelai

source