University of Michigan, Ann Arbor, MI
East Conference Room, Rackham Building

Thursday June 14, 2017

2:00-2:10pm Welcome and Introductory remarks

2:10-3:10pm Simon Haykin, “The Cognitive Dynamic System for Complex Networks with Risk under Control”

3:10-3:30pm coffee/tea break

3:30-4:10pm Yanhui Geng, “Network Mind: A Machine Learning-based Network Control Paradigm”
4:10-4:50pm Martha Steenstrup, “Adaptation and Learning in Network Control: Retrospective and Prospective“

4:50-5:00pm short break

4:50-5:30pm Group discussion: “Communications Networks as Playgrounds for Learning Research.”

Friday June 15, 2017

8:00-9:00am Breakfast

9:00-10:00am Xin Liu, “Learning Meets Networks: Challenges and Opportunities in Network Resource Allocation”

10:00-10:20am coffee/tea break

10:20-11:00am Karen Haigh, “Real-time Learning and Optimization for Wireless Communications Networks”
11:00-12:40am Group discussion: “Applying Learning Methods to Network Control Problems”.
Topics will include:
– is learning always a win for network control?
– which learning methods for which control problems and why?
– model-free vs. model-based learning?
– the role of deep learning
– overcoming reluctance to automate and decentralize

12:40-12:50pm Concluding remarks