giulia
Dr. Giulia Zarpellon
Polytechnique Montréal
Local AA-5505

After earning a master’s degree in Mathematics at University of Padova, Giulia is currently a Ph.D. candidate in Applied Mathematics at Polytechnique Montréal, under the supervision of prof. Andrea Lodi.

Giulia’s main research interests are in the integration of mathematical optimization and machine learning. Her doctoral research is focused on exploring some of the opportunities this synergy offers, with the aim of tightly integrating the learning process with the optimization one. Various are the decision-making situations that could benefit from a learning-based approach, especially in highly heuristic optimization frameworks such as the Mixed-Integer Programming one.

Giulia’s recent and current projects include a classification question on Mixed-Integer Quadratic Programs and a learning approach for tackling decisions in the branch-and-bound framework.

 

Articles

  • Bonami, P., Lodi, A. & Zarpellon, G. (2020) A Classifier to Decide on the Linearization of Mixed-Integer Quadratic Problems in CPLEX. IN Optimization Online preprint, ..
    [Bibtex]
    @article{BonamiLZ20,
    title={A Classifier to Decide on the Linearization of Mixed-Integer Quadratic Problems in {CPLEX}},
    author={Bonami, Pierre and Lodi, Andrea and Zarpellon, Giulia},
    journal={Optimization Online preprint},
    url={http://www.optimization-online.org/DB_HTML/2020/03/7662.html},
    year={2020}
    }
  • [DOI] Lodi, A. & Zarpellon, G. (2017) On learning and branching: a survey. IN TOP, 25.207–236.
    [Bibtex]
    @article{LodiZ17,
    author="Lodi, Andrea and Zarpellon, Giulia",
    title="On learning and branching: a survey",
    journal="TOP",
    year="2017",
    month="Jul",
    day="01",
    volume="25",
    number="2",
    pages="207--236",
    issn="1863-8279",
    doi="10.1007/s11750-017-0451-6"
    }

In Proceedings

  • Fischetti, M., Lodi, A. & Zarpellon, G. (2019) Learning MILP Resolution Outcomes Before Reaching Time-Limit IN Rousseau, L. & Stergiou, K. (Eds.), . Cham, Springer International Publishing, 275–291.
    [Bibtex]
    @InProceedings{FischettiLZ19,
    author="Fischetti, Martina and Lodi, Andrea and Zarpellon, Giulia",
    editor="Rousseau, Louis-Martin and Stergiou, Kostas",
    title="Learning {MILP} Resolution Outcomes Before Reaching Time-Limit",
    booktitle="Integration of Constraint Programming, Artificial Intelligence, and Operations Research",
    year="2019",
    publisher="Springer International Publishing",
    address="Cham",
    pages="275--291",
    }
  • Bonami, P., Lodi, A. & Zarpellon, G. (2018) Learning a Classification of Mixed-Integer Quadratic Programming Problems IN van Hoeve, W. (Ed.), .Springer International Publishing, 595–604.
    [Bibtex]
    @inproceedings{BonamiLZ18,
    author="Bonami, Pierre and Lodi, Andrea and Zarpellon, Giulia",
    editor="van Hoeve, Willem-Jan",
    title="Learning a Classification of Mixed-Integer Quadratic Programming Problems",
    booktitle="Integration of Constraint Programming, Artificial Intelligence, and Operations Research",
    year="2018",
    publisher="Springer International Publishing",
    pages="595--604",
    isbn="978-3-319-93031-2"
    }

Technical Reports

  • Zarpellon, G., Jo, J., Lodi, A. & Bengio, Y. (2020) Parameterizing Branch-and-Bound Search Trees to Learn Branching Policies .
    [Bibtex]
    @misc{ZarpellonJLB20,
    title={Parameterizing Branch-and-Bound Search Trees to Learn Branching Policies},
    author={Giulia Zarpellon and Jason Jo and Andrea Lodi and Yoshua Bengio},
    year={2020},
    eprint={2002.05120},
    archivePrefix={arXiv},
    primaryClass={cs.LG}
    }