Ph.D.'s degree in Process Mining
CentraleSupélec, Paris-Saclay - Gif-sur-Yvette, France
2022 - present
- Supervised by Paolo Ballarini and Pascale Le Gall
- MICS: Mathematics and Informatics
Pierre Cry
Ph.D. Graduate in Computer Science
Postdoctoral Researcher at INRAE
Keyword:
Process mining, Stochastic process discovery,
Petri net, Process tree, Bayesian inference.
Current situation
I am currently a Postdoctoral Researcher at INRAE Toulouse in the MIAT unit, working on probabilistic modeling and statistical inference for LiDAR-based forest characterization. My research focuses on developing Hidden Markov Random Field models to account for spatial dependencies in LiDAR observations and to estimate vegetation properties such as Leaf Area Density. I am also continuing my research activities following my Ph.D., with ongoing work on stochastic process discovery, probabilistic modeling, and statistical inference in process mining.
Research activities
My research lies within the field of process mining, with a particular focus on process discovery. I design algorithms and methods for discovering stochastic process models, such as stochastic Petri nets and process trees, that accurately capture real-world business processes' structural and probabilistic behaviours. My work bridges model inference, formal methods, and statistical techniques to support robust, interpretable, data-driven representations of complex systems.
My contributions include methods for extracting the stochastic behaviours of a model via reachability unfolding, optimization procedures guided by distance metrics between observed and simulated behaviours, and Bayesian inference approaches for learning probabilistic parameters from event logs. These methods are designed to be both data-driven and interpretable, providing robust tools for understanding variability and uncertainty in complex systems.