Trends in Recommendation & Personalization at Netflix

In this session (scl.ai/3EuVRw1), Justin Basilico discusses recent trends in personalization and the challenges of applying machine learning to recommendation systems to create better user experiences.
Executive Summary: Netflix Explains Recommendations & Personalization - scl.ai/3rDhwOZ
Full discussion transcript - scl.ai/3EuVRw1
Bio: Justin Basilico is a Director of Machine Learning and Recommender Systems at Netflix. He leads an applied research team that creates algorithms used to personalize the Netflix homepage through machine learning, recommender systems, and large-scale software engineering.
Join the leading AI community full of leaders, visionaries, practitioners, and researchers. Get full access to more discussions like this: exchange.scale.com

[@wonseokchung657]

Clearly presented. Helped a lot!

[@user-os7mk6bb5h]

Is it possible to access the presentation document (pptx / pdf)? In order to get the links to referred papers.

[@HazemAzim]

Very Insightful .. Thank you

[@andrewcampbell7011]

Any recommendations (haha) for papers about delayed reward in bandits and RL?

[@BryceChudomelka]

hah

[@podunkman2709]

I don't understand people who rave about Netflix's recommendations at all. This system is primitive. If you have watched comedies, it will suggest comedies to you. If your last movie was directed by X, the recommendation will be a movie made by X.
And the most common "recommendation" is what others watch.
This system NEVER offered me anything that made sense. It just doesn't work!