ServiceNow Research

Valentina Zantedeschi

Valentina Zantedeschi

Research Scientist

Human Decision Support

Valentina Zantedeschi is a Research Scientist in the Human Decision Support program at ServiceNow Research, and a member of the ELLIS Society. Her research focuses on building ML models that are inherently interpretable via latent discrete structures. My works span discrete optimization, structure learning, decentralized learning, learning theory, time-series analysis and applications to Earth science. Before joining ServiceNow, Valentina was a post-doctoral researcher at INRIA and University College London, in the context of the INRIA-London Programme, working in Benjamin Guedj’s team on PAC-Bayesian learning, in particular for voting classifiers. She holds a Ph.D. in Computer Science from Jean Monnet University (Saint-Étienne, France) and University of Lyon (France) with a focus on kernel learning with theoretical guarantees on performance, advised by Marc Sebban and Rémi Emonet. In 2017, she worked as a research intern at IBM Research, Dublin, in Mathieu Sinn’s team, studying and building Deep Learning architectures robust to adversarial examples. The library developed for this research work served as codebase for the release of Adversarial Robustness Toolbox.

  • Discrete Optimization
  • Learning Therory
  • Ensemble Methods
  • Causality
  • Time Series Forecasting


DAG Learning on the Permutahedron. International Conference of Learning Representations (ICLR),  2023.

PDF Cite Code

Discrete Learning Of DAGs Via Backpropagation. Workshop at the Neural Information Processing Systems (NeurIPS),  2022.

PDF Cite

On Margins and Generalisation for Voting Classifiers. Conference on Neural Information Processing Systems (NeurIPS),  2022.

PDF Cite