I am currently a post-doc within the Data Science for Real-Time Decision Making CERC (Canada Excellence Research Chair) at Polytechnique Montréal. I am interested in improving operational research methods using deep learning, under the supervision of Profs. Andrea Lodi and Laurent Charlin.
Research topics of interests
- Probabilistic Graphical Models: theoretical properties as conditional independence models (directed, undirected, bi-directed, chain, mixed graphs), structure learning algorithms.
- Multi-label classification: minimizing complex loss functions such as 0/1 error and F1 error.
- Deep learning: learn decomposable representations for structured prediction.