Contact information:

Sammy Khalife
Department of Applied Mathematics and Statistics
Johns Hopkins University
Wyman Park Building
3400 North Charles Street
Baltimore, MD 21218
Office: N418
Email: khalife.sammy[AT]jhu[DOT]edu

I'm a postdoctoral fellow in the Applied Mathematics and Statistics Department, at Johns Hopkins University, mentored by Amitabh Basu. My current research interests lie in the field of Discrete Optimization intersected with Theoretical Deep Learning. In particular, I am interested in the exact representability of classes of functions with Neural Networks, along with their computational complexity aspects (precise number of layers and neurons required), and the design of algorithms to reach global optimality for their training. I also work on other topics such as the formal expressiveness of Graph Neural Networks and the use of Deep Learning for Integer Programs.

  • Johns Hopkins University

    • Numerical Linear Algebra 553.385 (Spring 2022) Course link

    • Discrete Mathematics 553.171 (Fall 2021)

  • Ecole Polytechnique, France

    • Data Science Starter Program (2017-2019, Formation Big Data, EXED)

    • Machine learning INF554 (2019, Teaching Assistant)

  • Saint-Joseph University, Beirut

    • Master Data Science: Graph Theory, Probabilistic Graphical Models (Spring 2019, Instructor)

Research Papers

Preprints and work in progress
  • Khalife, S., Basu A. (2023). On the power of graph neural networks and the role of the activation function.

  • Khalife, S., Cheng, H., Basu A. (Extended version, 2022). Neural networks with linear threshold activations: structure and algorithms. (Mathematical Programming, in review)

  • Khalife, S., Basu A. (2022) Probabilistic cross-matching in random catalogues: optimal configurations.

  • Khalife, S., Ponty Y. Bulteau L. (Extended Version, 2021) Sequence graphs realizations and ambiguity in language models. (Journal of Theoretical Computer Science, in review).

  • Khalife, S., Achebouche R., Basile L., Virgile Dine, Myllykallio H. (2021). New learning architecture for protein ligand interaction.

Refereed Publications
  • Khalife, S., Basu A. (2022). Neural networks with linear threshold activations: structure and algorithms. In Integer Programming and Combinatorial Optimization (IPCO) 2022.

  • Khalife, S., Gonçalves, D., Allouah Y., and Liberti L. (2021). Further results on latent discourse models and word embeddings. Journal of Machine Learning Research, 22, pp. 1-36, November 2021.

  • Khalife, S., Gonçalves, D., and Liberti L. (2020). Distance geometry for word embeddings and applications. Journal of Computational Mathematics and Data Science.

  • Khalife, S. Ponty, Y., and Bulteau L. (2021). Sequence graphs realizations and ambiguity in language models. In International Computing and Combinatorics Conference (COCOON21)

  • Khalife, S., Malliavin T., and Liberti L. (2021). Secondary structure assignment of proteins in the absence of sequence information. Bioinformatics Advances, Volume 1, Issue 1, 2021.

  • Khalife, S. (2020). Sequence graphs: characterization and counting of admissible elements. Cologne-Twente Workshop on Graphs and Combinatorial Optimization.

  • Khalife, S., Read, J., and Vazirgiannis, M. (2020). Structure and influence analysis of worldwide capitalistic ownership. Journal of Applied Network Science.

  • Khalife, S., Liberti, L., and Vazirgiannis, M. (2019a). Geometry and analogies: a study and propagation method for word representations. In International Conference on Statistical Language and Speech Processing, pages 100–111. Springer

  • Khalife, S. and Vazirgiannis, M. (2019). Scalable graph-based method for individual named entity identification. In Proceedings of the Thirteenth Workshop on Graph-Based Methods for Natural Language Processing (TextGraphs-13), pages 17–25

  • Khalife, S., Read, J., and Vazirgiannis, M. (2019b). Empirical analysis of a global capital-ownership network. In International Conference on Complex Networks and Their Applications, pages 656–667. Springer

  • Pallanca, O., Khalife, S., and Read, J. (2018). Detection of sleep spindles in NREM 2 sleep stages: Preliminary study & benchmarking of algorithms. In Zheng, H. J., Callejas, Z., Griol, D., Wang, H., Hu, X., Schmidt, H. H. H. W., Baumbach, J., Dickerson, J., and Zhang, L., editors, IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2018, Madrid, Spain, December 3-6, 2018, pages 2652–2655. IEEE Computer Society


  • Thèse (Ecole Polytechnique): Graphs, Geometry and Representations for Language Models and Networks of Entities
    Advisor: Michalis Vazirgiannis