RMP practice webinar series

Speaker: Nicolas Stier-Moses, Director at Facebook Core Data Science
Title: Operations Management Research at Facebook
Time: 8-9 pm Eastern Time (5-6 pm Pacific Time) on Wednesday May 12th, 2021
Moderator: Daniela Saban, Stanford University

Registration form
Bio: Dr. Nicolas Stier-Moses
Nicolas Stier-Moses is a Director at Facebook Core Data Science. His work leverages innovative research to drive impact to the products, infrastructure and processes at Facebook, the company. The group draws inspiration from a rich and diverse set of disciplines including Operations, Economics, Mechanism Design, Algorithms, Statistics, Machine Learning, Experimentation, and Computational Social Science. Between 2014 and 2017, he supported the Economics, Algorithms and Optimization team, which is one of the areas of focus of Core Data Science. Prior to joining Facebook, Nicolas was an Associate Professor at the Decision, Risk and Operations Division of Columbia Business School and at the Business School of Universidad Torcuato Di Tella. He received a Ph.D. degree from the Operations Research Center at the Massachusetts Institute of Technology.
Abstract:
Operations Management research at Facebook

In this webinar I'll give a high-level overview a few research directions in the area of Operations Management that are used to derive operational insights and product improvement at Facebook. Related to marketplace modeling, which found widespread use in high tech firms and is central to the RMP community, I'll discuss pacing equilibria which capture the alignment between budgets and bids in sequential auction settings. I'll also cover the basics of Bayesian Optimization and its relation to value modeling, which has had widespread adoption in ranking of feed surfaces. Finally, I will present some work related to large-scale detection systems for violating content in platforms. This includes estimating virality, training classifiers, trade-offs between accuracy and latency when performing multiple reviews for each label, as well as queuing systems for capacity planning, prioritization, and admission control.

Some material related to the content to be presented:

Pacing equilibria, second price auctions, WINE 2018, OR 2021: https://arxiv.org/pdf/1706.07151.pdf
Pacing equilibria, first price auctions, EC 2019: https://arxiv.org/pdf/1811.07166.pdf
Equilibrium abstractions, EC 2019, OR 2021: https://arxiv.org/pdf/1901.06230.pdf
Bayesian optimization for experimentation and value model tuning (several papers): https://botorch.org/
Estimating spread in social networks: https://arxiv.org/abs/2009.02092
Temporal embeddings, KDD 2020: https://research.fb.com/wp-content/uploads/2020/08/TIES-Temporal-Interaction-Embeddings-For-Enhancing-Social-Media-Integrity-At-Facebook.pdf
Multiple reviews for detecting violating content, KDD 2020: https://research.fb.com/wp-content/uploads/2020/08/CLARA-Confidence-of-Labels-and-Raters.pdf
Queueing for content moderation at scale: https://arxiv.org/abs/2103.16816
Name *
Email *
Company or University Name *
Job Position Title *
Do you have any specific question for Nicolas?
Submit
Never submit passwords through Google Forms.
This form was created inside of University of Chicago, Booth School of Business. Report Abuse