Programme


First day: May 24, 2018

8:30am - 9:10am       Registration and Coffee

9:10am - 9:15am       Gui Liberali and Vianney Perchet - Welcome Note

9:15am -10:00am      Warren Powell (Princeton): From Multi-armed Bandits to Stochastic Optimization. For tutorial see here

10:00am -10:30am    Clément Bouttier  (Airbus, Toulouse): Piyavskii's algorithm for deterministic or stochastic Lipschitz global optimization

10:30am -11:00am    Nicolas Grislin (AlephD, Paris): Optimization of a SSP's Header Bidding Strategy using Thompson Sampling

11:00am -11:30pm    Coffee Break

11:30am -12:00pm    Daria Dzyabura (New York University):  Optimizing for path-dependence in consumer cross-category search

12:00am -12:30pm    Song Lin (HKUST): Informational Complementarity

12:30am -1:00pm      Shervin Tehrani (Toronto and UTDallas): A heuristic approach to explore: value of perfect information

1:00pm -2:30pm        Lunch and Poster Session

2:30pm  - 3:15pm      Nicolò Cesa-Bianchi (Dipartimento di Informatica, Università degli Studi di Milano): NonStochastic Bandits with Anonymous Feedback.

3:15pm - 4:45pm       Pierre Alquier (Ensae): Regret Bounds for Lifelong Learning      

3:45pm - 4:15pm      Claire Vernade (Amazon):   Stochastic Bandit Models for Delayed Conversions     

4:15pm - 4:45pm      Richard Combes (Centrale Supelec): Minimal Exploration in Structured Stochastic Bandits

4:45pm - 5:15pm       Coffee Break

5:15pm - 5:45pm       Trivikram Dokka (Lancaster University): Non-parametric robust dynamic pricing: a non-adversarial approach

5:45pm  - 6:15pm      Arnoud den Boer (University of Amsterdam): Discontinuous demand functions: estimation and pricing.  Working paper

6:15pm  - 6:45pm      Jan Svitak (Dutch Authority for Consumer & Markets) 
                                       Mallesh Pai (Rice University).  Can Competing Bandit Algorithms Lead to Non-Competitive Prices?

6:45pm - 9:00pm        Dinner at In Den Rustwat

 

Second day: May 25, 2018

9:00am -9:45am           Miguel Villas-Boas   (University of California, Berkeley): Optimal Learning before Choice

9:45am -10:15am        Quentin Berthet (Cambridge): Fast Rates for Bandit Optimization with Upper-Confidence Frank-Wolfe

10:15am -10:45am       Siham el Kihal (Frankfurt School of Finance & Management): Product Return Management in Omnichannel Retail

10:45am -11:15pm       Coffee Break

11:15am -11:45pm       Xiao Liu (New York University): The Effect of Word of Mouth on Sales: New Answers from the Consumer Journey Data with Deep Learning

11:45am -12:15pm       Lan Luo (University of Southern California): Can User Generated Content Predict Restaurant Survival? Deep Learning of Yelp Photos and Reviews

12:15pm -1:30pm         Lunch 

1:30pm - 2:15pm          Donald A. Berry (U. Texas MD Anderson Cancer Center): Multi-armed Bayesian bandits Change the Paradigm of Clinical Trials

2:15pm - 2:45pm          Sofia Villar (Cambridge): Developing Implementable Bandit-Based Designs for Clinical Trials: Where Methods Meet Practice

2:45pm - 3:15pm          Peter Jacko (Lancaster University): Mitigating the Curse of Dimensionality of the Bayesian Beta-Bernoulli Bandit Problem

3:15pm - 4:00pm          Coffee Break and Transfer to the Erasmus Building - Aula Auditorium

4:00pm - 4:45pm          Gui Liberali (Erasmus University): Inaugural Address on (Machine) Learning with a Purpose . Pdf available here

4:45pm     Reception