William Trouleau

William Trouleau

Engagement Lead & Data Scientist

Unit8

Hi! I’m William!

I’m an Engagement Lead & Data Scientist at Unit8 where I use data and machine learning to solve hands-on business challenges. I work on the whole journey from design, to implementation and deployment of AI-based systems for customers in the chemical industry spanning different business units: from innovation, to compliance or manufacturing.

In 2021, I earned a PhD from EPFL in Lausanne, Switzerland. My research was focused on the statistical and algorithmic aspects of modeling, control, and inference of networks of times series ; with applications in epidemiology, neuroscience, information diffusion and recommendation systems.

During my PhD, I was an intern at the Institute for Disease Modeling in Seattle, quantifying the burden of Tuberculosis in Pakistan. I also had the opportunity to intern at Technicolor in the Silicon Valley.

Activities
  • ๐Ÿš€ Joined Unit8 as Data Scientist (2021)
  • ๐ŸŽ“ Defended my PhD thesis (2021)
  • ๐Ÿฆ  Winner of the KIT Tuberculosis Hackathon as part of the IDM team (2019)
  • ๐Ÿ‘จโ€๐Ÿซ Received Teaching Assistant Award, EPFL (2019)
  • โœˆ๏ธ Received ACM SIGKDD Student Travel Award (2016)
  • ๐Ÿ“ Awarded EDIC Fellowship, EPFL (2015)

Publications

(2022). Quantifying the Effects of Contact Tracing, Testing, and Containment Measures in the Presence of Infection Hotspots. ACM Transactions on Spatial Algorithms and Systems.

PDF Code

(2021). Learning Self-Exciting Temporal Point Processes Under Noisy Observations. PhD thesis, EPFL.

PDF

(2021). A Variational Inference Approach to Learning Multivariate Wold Processes. Proceedings of the 24th International Conference on Artificial Intelligence and Statistics (AISTATS).

PDF Code

(2019). Learning Hawkes Processes from a Handful of Events. Proceedings of the 36th International Conference on Machine Learning (ICML).

PDF Code

(2019). Learning Hawkes Processes Under Synchronization Noise. Proceedings of the 36th International Conference on Machine Learning (ICML).

PDF Code

(2018). Stochastic Optimal Control of Epidemic Processes in Networks. Machine Learning for Health (ML4H) Workshop at NeurIPS.

PDF Code

(2018). Preventive chemotherapy to control soil-transmitted helminthiasis averted more than 500 000 DALYs in 2015. Proceedings of the 24th International Conference on Artificial Intelligence and Statistics (AISTATS).

Full Text

(2016). Just one more: Modeling binge watching behavior. Proceedings of the 22nd International Conference on Knowledge Discovery and Data Mining (KDD).

PDF