Program

The workshop takes place on Friday 13th of September, in room Zeta 2, starting at 2pm.

In each section we will be referring to relevant literature, as well as pointing out open research directions with clearly outlined potential novel contributions. We hope this would stimulate research community to pursue the solutions to the outlined problems.

Part 1: Experimentation & Causal Inference in industry

• How experiments are used from product development
• What kind of experiment designs are used
• What are the examples of hypotheses tested
• What statistical challenges do these experiments face
• What metrics are measured
• Aggregating experiment impact
• Any-time valid inference
• Treatment effects distribution

Part 2: Examples of statistical challenges and research opportunities

• Peeking
• Variance reduction techniques using machine learning

Part 3: Causal inference in two-sided marketplace setting

• Challenges of experimentation in two sided marketplaces
• Measuring effects in the opposite side of the marketplace: bipartite graphs and machine learnings