Mélanie MUNCH

I am currently a researcher in the Protein, Structure, Function team at INRAE STLO (Science and Technology of Milk and Eggs). My research focuses on the application of machine learning techniques for modelling complex systems and multi-criteria decision making. I am interested in the interactions between expert knowledge and modelling approaches. After a degree in agricultural engineering from AgroParisTech, I prepared my PhD on the combination of probabilistic relational models and knowledge bases, with the aim of highlighting the importance of integrating expert knowledge in automatic methods for a better explainability of the learned models. Since 2022, I am also co-moderator of the INRAE MathNum scientific network [IN-OVIVE] (https://www6.inrae.fr/reseau-in-ovive/). Some of my work is presented in my [Git-Lab](https://gitlab.com/melanie.munch). ### Competences Artificial intelligence and machine learning algorithms with an emphasis on model explainability Expert knowledge integration Causality and causal discovery Complex system modeling ### Positions **Jan 2023** Research Engineer. UMR STLO, PSF team, INRAE Rennes, France. **Jan 2021 - Dec 2023** Postdoctoral fellow. I2M Team, Bordeaux University, France. *Wheat Quality Modeling* **Oct 2020 - Dec 2021** Research engineer. UMR IATE, Ico Team, INRAE Centre Occitanie-Montpellier, France. *Eco-design of bio-composite packaging by combining probabilistic graphical models and knowledge graphs* and *Prediction of solubility of foods as a function of composition* **Sep 2019 - Dec 2019** Visiting researcher. DISI Team, Trento University, Italy. *Learning under constraint for choice preference modeling* ### Academic Training **2017- 2020** PhD in applied informatics. *Improving uncertain reasoning combining probabilistic relational models and expert knowledge* (AgroParisTech, France). **2017** Master of Science: Computer Science, Intelligent Systems (Paris Dauphine University, France) **2013-2017** Engineer Diploma (AgroParisTech, France). ### Oral Communications **ICBC 24** Smart use of sensory data for wheat grain quality grading **PangBorn 15** Combining Probabilistic Models With Expert Knowledge Integration For Wheat Quality Assessment **ICEF 14** Expert knowlege integration for smart tools modeling: application to wheat use ### Teachings Python and Database, L3 level students (AgroParisTech) Excel VBA, M2 level students (AgroParisTech) ### Grants and Awards **Fait Marquant INRAe - Département TRansform 2022 :** [Prediction of CO2 solubility in foods for modified atmosphere packaging](https://www.inrae.fr/actualites/rapport-recherche-innovation-du-departement-transform-2023-est-paru) **Fait Marquant INRAe - Département TRansform 2021 :** [Modeling of bio-composite packaging by integration of data and expert knowledge ](https://www.inrae.fr/actualites/rapport-recherche-innovation-du-departement-transform-2022-est-paru) **Best research paper:** Melanie Munch, Patrice Buche, Cristina Manfredotti, Pierre-Henri Wuillemin, Helene Angellier-Coussy. A process reverse engineering approach using Process and Observation Ontology and Probabilistic Relational Models: application to processing of bio-composites for food packaging. *15th International Conference on Metadata and Semantics Research*, Nov 2021, Madrid, Spain. **2019 Dufrenoy Grant** from the agricultural academy for innovative science project **EIR-A label** from the Agreenium international research center for an interdisciplinary and international opening during the PhD # Others Contributor to the [PyAgrum](https://agrum.gitlab.io/) library Printemps de la donnée (INRAE): <https://nextcloud.inrae.fr/s/MPXoktJRYDYpTLd> (french) ### Reviews CARI2022, EGC2023, Journal of Expert Systems with Application