Schedule - UTC TIME ZONE

TO ATTEND THE VIRTUAL WORKSHOP PLEASE GO TO https://icml.cc/virtual/2020/workshop/5738


FIRST SESSION

12:00 PM UTC -- Introduction and opening remarks. Workshop organizers.

12:10 PM UTC -- Invited talk: "Deploying Machine Learning Models in a Developing World Context", Ernest Mwebaze

12:55 PM UTC -- Invited talk: "System-wide Monitoring Architectures with Explanations ", Leilani Gilpin


01:40 PM UTC - 01:50 PM UTC-- BREAK


01:50 PM UTC -- Invited talk: "Bridging the gap between research and production in machine learning", Chip Huyen

02:20 PM UTC - 03:40 PM UTC -- Contributed talks

      • 02:20 PM UTC -- "Monitoring and explainability of models in production", Klaise

      • 02:30 PM UTC -- "Gradient-Based Monitoring of Learning Machines", Liu

      • 02:40 PM UTC -- "Not Your Grandfather's Test Set: Reducing Labeling Effort for Testing", Taskazan

      • 02:50 PM UTC -- "Carbontracker: Tracking and Predicting the Carbon Footprint of Training Deep Learning Models", Anthony

      • 03:00 PM UTC -- "Serverless inferencing on Kubernetes", Cox

      • 03:10 PM UTC -- "Do You Sign Your Model?", Aramoon

      • 03:20 PM UTC -- "PareCO: Pareto-aware Channel Optimization for Slimmable Neural Networks", Chin

      • 03:30 PM UTC -- "Technology Readiness Levels for Machine Learning Systems ", Lavin

03:40 PM UTC -- Virtual Poster session


04.20 PM UTC - 4.30 PM UTC -- BREAK


SECOND SESSION

4:30 PM UTC -- Open problems: panel discussion

              • Chairs: Alessandra Tosi, Nathan Korda

              • Panelists (problem descriptions available here):
                Zhenwen Dai (Spotify): Model Selection for Production Systems
                Erick Galinkin (Montreal AI Ethics Institute and Rapid7): Green Lighting ML
                Camylle Lanteigne (MAIEI and McGill University): SECure: A Social and Environmental Certificate for AI Systems
                Deploy machine learning models serverlessly at scale
                Yuzhui Liu (Bloomberg): Deploy machine learning models serverlessly at scale
                Alexander Lavin                 (Augustus Intelligence): ML lacks the formal processes and industry standards of other engineering disciplines.
                Alexander Lavin (Augustus Intelligence): Approaches to AI ethics must consider second-order effects and downstream uses, but how?

05:30 PM UTC - 06:00 PM UTC -- BREAK


6:00 PM UTC -- Invited talk: "Successful Data Science in Production Systems: It’s All About Assumptions", Nevena Lalic

6:45 PM UTC -- Invited talk: "Conservative Exploration in Bandits and Reinforcement Learning", Mohammad Ghavamzadeh

7:30 PM UTC -- Panel discussion

              • Chair: TBA

              • Panelists: Ernest Mwebaze, Leilani Gilpin, Nevena Lalic, Mohammad Ghavamzadeh.

8:30 PM UTC -- Final remarks. Workshop organizers.

8:35 PM UTC -- End of the day