Position-Based Multiple-Play Bandits with Thompson Sampling
Gauthier, C.-S.; Gaudel, R.; Fromont, E. Position-based multiple-play bandits with thompson sampling. IDA’21, 2021.
CV
Work experience
- Since March 2019: PhD Student/Research Scientist
- Digital innovation, Louis Vuitton
- Duties included: Recommendation Systems
- Supervisor: Romaric Gaudel, Elisa Fromon, Bruno Guilbot & Eliot Barril
- Sept 2018 - Feb 2019: Research Scientist
- Digital innovation, Louis Vuitton
- Duties included: Recommendation Systems, Navigation Analysis
- Manager: Bruno Guilbot
- April 2018 - Sept 2018: Data Scientist Intern
- Lv_Lab, Louis Vuitton
- Duties included: Personnalization of LV.com’s search engine
- Supervisor: Bruno Guilbot
- Summer 2017: Research Intern
- Scoring center, BNP Paribas Personnal Finance
- Duties included: Inclusion of high cardinality variables’ exploration
- Supervisor: Olga Fazylova
- Summer 2016: Intern
- Royal British Society of Scultpture
- Duties included: Touring, Archiving
- Supervisor: Marcus O’Neil
Education
- Ph.D in Computer Sciences, subject: Recommendation system and Bandits Algorithm, University of Rennes 1, 2022 (expected)
- Engineer’s degree (equivalent to a master’s degree) in Quantitative Marketing and Revenue Manadgement, ENSAI, 2018
- Bachelor’s degree (double degree with ENSAI cursus) in Mathematical engineering, University of Rennes 1, 2016
- Classe preparatoire MPSI/MP, Lycée Saint Louis, Paris, 2015
Skills
- Machine Learning
- clustering
- outlier detection
- scoring
- Bandits
- Statistical Analysis
Publications
Parametric Graph for Unimodal Ranking Bandit
Gauthier, C.-S.; Gaudel, R.; Fromont, E.; and Lompo, B. A. 2021. Parametric Graph for Unimodal Ranking Bandit. ICML’21,2021.
Talks
Joint Conferences CAp and RFIAP 2020
talk at Joint Conferences CAp and RFIAP 2020, Vannes (virtual), France
MT180
talk at Three-minute Thesis, Rennes, France
Conference CAp 2021
talk at Conference CAp 2021, Saint Etienne (virtual), France
Teaching & Supervision
- Recommendation System : Pratical session (12h)
- Neural Network 2019: Pratical session (6h)
- Intership on Contextual Bandit: Supervision of Léa Fortat (6 month)
Project and leadership
- Project supervision
- title: “Contextual dynamic recommendation system for Louis Vuitton’s website”
- type: Final Project, supervision of three students during 5 month to help them apply what they learned on an industrial problematic
- Multiple Hackathon within Louis Vuitton and LVMH ecosystem
- Various subject: supply chain, data, CRM, media …
- example: DARE
- Active member of Ensai Junior Consultant : stundent association
- 2015: head of internal control
- 2016: vice president
- Project: Classification and forcast of days according to the number of hotel reservation, Disneyland Resord Paris, 2018
- Project: Forcast of buses’consumptions, Transdev,2017
- Data challenge: Optimisation of client satisfaction in railway, SNCF,2016
- Project: Analysis of scoial and economical situation of artists affiliated to MDA, Culture Ministry, 2016