Daniele Zambon

I am a post-doc at the Swiss AI Lab IDSIA and USI (CH) in the graph machine learning group.

The focus of my research is graph representation learning, learning in non-stationary environments, and graph stream processing.

firstname.surname@usi.ch
GitHub  ·  Google Scholar  ·  LinkedIn  ·  ORCID

News!

• Paper accepted @ NeurIPS ‘22. AZ-whiteness test: with it you can test the optimality of GNN’s.
• We got funded by OCRE with cloud computing resources for a project in collaboration with UniPD.

About me

I obtained my Ph.D. from USI Università della Svizzera italiana (CH) under the supervision of Prof. Cesare Alippi (from USI and POLIMI) and Prof. Lorenzo Livi (from Univ. of Manitoba, CA, and Univ. of Exeter, UK). My research dealt with statistical tests for anomaly and change detection in sequences of graphs, graph representation learning, and learning in non-stationary environments. You can find a list of my publications here.

I have been visiting researcher at the University of Florida (US) working on kernel adaptive methods and at the University of Exeter (UK) exploring embeddings onto Riemannian manifolds. I have also been an intern at STMicroelectronics (Italy) where I developed my Master’s thesis on sparse models for anomaly detection. I received Master’s and Bachelor’s degrees from the Università degli Studi di Milano (Italy) focusing on approximation theory and mathematical statistics.

I am (or have been) in the program committee of top-tier conferences and journals of the field, including IEEE TNNLS, IEEE TSP, IEEE PAMI, IEEE IJCNN, NeurIPS, ICLR, ICML, CVPR.

Publications

Preprints

• Graph Kalman Filters, 2023, Alippi*, Zambon*.
  [ arXiv] [bibtex]

  @misc{alippi2023graph, title = {Graph {{Kalman Filters}}}, author = {Alippi, Cesare and Zambon, Daniele}, year = {2023}, month = mar, number = {arXiv:2303.12021}, publisher = {{arXiv}}, doi = {10.48550/arXiv.2303.12021}, archiveprefix = {arxiv}}

• Taming Local Effects in Graph-based Spatiotemporal Forecasting, 2023, Cini*, Marisca*, Zambon, Alippi.
  [ arXiv] [bibtex]

  @misc{cini2023taming, doi={10.48550/ARXIV.2302.04071}, url={https://arxiv.org/abs/2302.04071}, author={Cini, Andrea and Marisca, Ivan and Zambon, Daniele and Alippi, Cesare}, keywords={Machine Learning (cs.LG), Artificial Intelligence (cs.AI), FOS: Computer and information sciences, FOS: Computer and information sciences}, title={Taming Local Effects in Graph-based Spatiotemporal Forecasting}, publisher={arXiv}, year={2023}, copyright={arXiv.org perpetual, non-exclusive license} }

• Where and How to Improve Graph-based Spatio-Temporal Predictors, 2023, Zambon, Alippi.
  [ arXiv] [bibtex]

  @misc{zambon2023where, doi={10.48550/ARXIV.2302.01701}, url={https://arxiv.org/abs/2302.01701}, author={Zambon, Daniele and Alippi, Cesare}, keywords={Machine Learning (stat.ML), Machine Learning (cs.LG), FOS: Computer and information sciences, FOS: Computer and information sciences}, title={Where and How to Improve Graph-based Spatio-Temporal Predictors}, publisher={arXiv}, year={2023}, copyright={arXiv.org perpetual, non-exclusive license} }

• Graph state-space models, 2023, Zambon, Cini, Livi, Alippi.
  [ arXiv] [bibtex]

  @misc{zambon2023graph, doi={10.48550/ARXIV.2301.01741}, url={https://arxiv.org/abs/2301.01741}, author={Zambon, Daniele and Cini, Andrea and Livi, Lorenzo and Alippi, Cesare}, keywords={Machine Learning (cs.LG), FOS: Computer and information sciences, FOS: Computer and information sciences}, title={Graph state-space models}, publisher={arXiv}, year={2023}, copyright={arXiv.org perpetual, non-exclusive license} }

• Sparse Graph Learning for Spatiotemporal Time Series, 2022, Cini, Zambon, Alippi.
  [ arXiv] [bibtex]

  @misc{cini2022sparse, doi={10.48550/ARXIV.2205.13492}, url={https://arxiv.org/abs/2205.13492}, author={Cini, Andrea and Zambon, Daniele and Alippi, Cesare}, keywords={Machine Learning (cs.LG), Artificial Intelligence (cs.AI), FOS: Computer and information sciences, FOS: Computer and information sciences}, title={Sparse Graph Learning for Spatiotemporal Time Series}, publisher={arXiv}, year={2022}, copyright={Creative Commons Attribution 4.0 International} }

Journals

• Understanding Pooling in Graph Neural Networks, IEEE TNNLS, 2022, Grattarola, Zambon, Bianchi, Alippi.
  [ DOI]  [ arXiv] [ GitHub] [bibtex]

  @article{grattarola2022understanding, title={Understanding Pooling in Graph Neural Networks}, author={Grattarola, Daniele and Zambon, Daniele and Bianchi, Filippo and Alippi, Cesare}, journal={IEEE Transactions on Neural Networks and Learning Systems}, year={2022}, volume={}, number={}, pages={1-11}, doi={10.1109/TNNLS.2022.3190922} }

• Change-Point Methods on a Sequence of Graphs, IEEE TSP, 2019, Zambon, Alippi, Livi.
  [ DOI]  [ arXiv] [ GitHub] [bibtex]

  @article{zambon2019change, title={Change-Point Methods on a Sequence of Graphs}, author={Zambon, Daniele and Alippi, Cesare and Livi, Lorenzo}, year={2019}, doi={10.1109/TSP.2019.2953596}, journal={IEEE Transactions on Signal Processing}, }

• Change Detection in Graph Streams by Learning Graph Embeddings on Constant-Curvature Manifolds, IEEE TNNLS, 2019, Grattarola, Zambon, Alippi, Livi.
  [ DOI]  [ arXiv] [ GitHub] [bibtex]

  @article{grattarola2019change, title={Change Detection in Graph Streams by Learning Graph Embeddings on Constant-Curvature Manifolds}, author={Grattarola, Daniele and Zambon, Daniele and Alippi, Cesare and Livi, Lorenzo}, year={2019}, doi={10.1109/TNNLS.2019.2927301}, journal={IEEE Transactions on Neural Networks and Learning Systems}, }

• Concept Drift and Anomaly Detection in Graph Streams, IEEE TNNLS, 2018, Zambon, Alippi, Livi.
  [ DOI]  [ arXiv] [experiments] [bibtex]

  @article{zambon2018concept, title={Concept Drift and Anomaly Detection in Graph Streams}, author={Zambon, Daniele and Alippi, Cesare and Livi, Lorenzo}, year={2018}, doi={10.1109/TNNLS.2018.2804443}, journal={IEEE Transactions on Neural Networks and Learning Systems}, pages={1-14}, }

Conferences

• AZ-whiteness test: a test for signal uncorrelation on spatio-temporal graphs, NeurIPS, 2022, Zambon, Alippi.
  [ arXiv] [ GitHub] [bibtex]

  @inproceedings{ zambon2022aztest, title={{AZ}-whiteness test: a test for signal uncorrelation on spatio-temporal graphs}, author={Daniele Zambon and Cesare Alippi}, booktitle={Advances in Neural Information Processing Systems}, editor={Alice H. Oh and Alekh Agarwal and Danielle Belgrave and Kyunghyun Cho}, year={2022}, url={https://openreview.net/forum?id=SFeKNSxect} }

• Deep Learning for Graphs, ESANN, 2022, Bacciu, Errica, Navarin, Pasa, Zambon.
  [ DOI]  [bibtex]

  @inproceedings{bacciu2022deep, title={Deep Learning for Graphs}, author={Bacciu, Davide and Errica, Federico and Navarin, Nicol\`o and Pasa, Luca and Zambon, Daniele}, booktitle={30th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, ESANN 2019}, year={2022}, organization={ESANN (i6doc. com)} }

• Graph iForest: Isolation of anomalous and outlier graphs, IEEE IJCNN, 2022, Zambon, Livi, Alippi.
  [ DOI]  [ GitHub] [bibtex]

  @inproceedings{zambon2022graph, title={Graph {iForest}: Isolation of anomalous and outlier graphs}, author={Zambon, Daniele and Livi, Lorenzo and Alippi, Cesare}, booktitle={2022 International Joint Conference on Neural Networks (IJCNN)}, pages={1--8}, year={2022}, organization={IEEE} }

• Understanding Catastrophic Forgetting of Gated Linear Networks in Continual Learning, IEEE IJCNN, 2022, Munari, Pasa, Zambon, Alippi, Navarin.
  [ DOI]  [ GitHub] [bibtex]

  @inproceedings{munari2022understanding, title={Understanding Catastrophic Forgetting of Gated Linear Networks in Continual Learning}, author={Munari, Matteo and Pasa, Luca and Zambon, Daniele and Alippi, Cesare and Navarin, Nicol\`o}, booktitle={2022 International Joint Conference on Neural Networks (IJCNN)}, pages={1--8}, year={2022}, organization={IEEE} }

• Erkennung von Anomalien und Veränderung in Graphsequenzen, D22, 2022, Zambon.
  [bibtex]

  @inproceedings{zambon2022erkennung, author = {Zambon, Daniele}, title = {Erkennung von Anomalien und Veränderung in Graphsequenzen}, booktitle={D22}, year={2022}, editor={Hölldobler, Steffen}, pages={311-320}, publisher={Köllen Druck + Verlag GmbH}, address={Bonn} }

• Graph Edit Networks, ICLR, 2021, Paassen, Grattarola, Zambon, Alippi, Hammer.
  [ GitLab] [bibtex]

  @inproceedings{paassen2021graph, title={Graph Edit Networks}, author={Paassen, Benjamin and Grattarola, Daniele and Zambon, Daniele and Alippi, Cesare and Hammer, Barbara Eva}, booktitle={International Conference on Learning Representations (ICLR)}, year={2021}, url={https://openreview.net/forum?id=dlEJsyHGeaL} }

• Graph Random Neural Features for Distance-Preserving Graph Representations, ICML, 2020, Zambon, Alippi, Livi.
  [ arXiv] [ GitHub] [bibtex]

  @inproceedings{zambon2020graph, title={Graph Random Neural Features for Distance-Preserving Graph Representations}, author={Zambon, Daniele and Alippi, Cesare and Livi, Lorenzo}, booktitle={Proceedings of the 37th International Conference on Machine Learning (ICML)}, pages={10968--10977}, editor={Hal Daumé III and Aarti Singh}, volume={119}, series={Proceedings of Machine Learning Research}, address={Virtual}, year={2020}, month={13--18 Jul}, publisher={PMLR}, pdf={http://proceedings.mlr.press/v119/zambon20a/zambon20a.pdf}, url={http://proceedings.mlr.press/v119/zambon20a.html}, abstract={We present Graph Random Neural Features (GRNF), a novel embedding method from graph-structured data to real vectors based on a family of graph neural networks. The embedding naturally deals with graph isomorphism and preserves the metric structure of the graph domain, in probability. In addition to being an explicit embedding method, it also allows us to efficiently and effectively approximate graph metric distances (as well as complete kernel functions); a criterion to select the embedding dimension trading off the approximation accuracy with the computational cost is also provided. GRNF can be used within traditional processing methods or as a training-free input layer of a graph neural network. The theoretical guarantees that accompany GRNF ensure that the considered graph distance is metric, hence allowing to distinguish any pair of non-isomorphic graphs.}}

• Graph Embeddings from Random Neural Features, GRL @ NeurIPS, 2019, Zambon, Alippi, Livi.
  [ GitHub] [bibtex]

  @inproceedings{zambon2019graph, title={Graph Embeddings from Random Neural Features}, author={Zambon, Daniele and Alippi, Cesare and Livi, Lorenzo}, year={2019}, booktitle={Advances in Neural Information Processing System (NeurIPS), Graph Representation Learning Workshop}, }

• Autoregressive Models for Sequences of Graphs, IEEE IJCNN, 2019, Zambon, Grattarola, Alippi, Livi.
  [ DOI]  [ arXiv] [ GitHub] [bibtex]

  @inproceedings{zambon2019autoregressive, title={Autoregressive Models for Sequences of Graphs}, author={Zambon, Daniele and Grattarola, Daniele and Alippi, Cesare and Livi, Lorenzo}, year={2019}, doi={10.1109/IJCNN.2019.8852131}, booktitle={IEEE International Joint Conference on Neural Networks}, }

• Anomaly and Change Detection in Graph Streams through Constant-Curvature Manifold Embeddings, IEEE IJCNN, 2018, Zambon, Livi, Alippi.
  [ DOI]  [ arXiv] [experiments] [bibtex]

  @inproceedings{zambon2018anomaly, title={Anomaly and Change Detection in Graph Streams through Constant-Curvature Manifold Embeddings}, author={Zambon, Daniele and Livi, Lorenzo and Alippi, Cesare}, year={2018}, doi={10.1109/IJCNN.2018.8489762}, booktitle={IEEE International Joint Conference on Neural Networks}, }

• Detecting Changes in Sequences of Attributed Graphs, IEEE SSCI, 2017, Zambon, Livi, Alippi.
  [ DOI]  [experiments] [bibtex]

  @inproceedings{zambon2017detecting, title={Detecting Changes in Sequences of Attributed Graphs}, author={Zambon, Daniele and Livi, Lorenzo and Alippi, Cesare}, year={2017}, doi={10.1109/SSCI.2017.8285273}, booktitle={IEEE Symposium Series on Computational Intelligence}, }

• Ecg monitoring in wearable devices by sparse models, ECML PKDD, 2016, Carrera, Rossi, Zambon, Fragneto, Boracchi.
  [ DOI]  [bibtex]

  @inproceedings{carrera2016ecg, title={Ecg monitoring in wearable devices by sparse models}, author={Carrera, Diego and Rossi, Beatrice and Zambon, Daniele and Fragneto, Pasqualina and Boracchi, Giacomo}, booktitle={Joint European Conference on Machine Learning and Knowledge Discovery in Databases}, pages={145--160}, year={2016}, organization={Springer} }

Thesis

• Anomaly and Change Detection in Sequences of Graphs, PhD Thesis, 2022, Zambon.
  [bibtex]

  @phdthesis{zambon2022phdthesis, title={Anomaly and Change Detection in Sequences of Graphs}, school={Universit{\`a} della Svizzera italiana}, publisher={Universit{\`a} della Svizzera italiana}, author={Zambon, Daniele}, year={2022}, }

Teaching

I regularly serve as a teacher in Master’s and Bachelor’s degree programs at USI holding lectures, lab sessions, and examinations.

• Machine Learning for Medical Students (Fall 2022-2023, Spring 2022-2023).
• Graph Deep Learning (Spring 2020-2021, Spring 2021-2022, Spring 2022-2023).
• Advanced Topics in Machine Learning (Fall 2020-2021).
• Machine Learning (Spring 2018-2019, Spring 2019-2020).
• Probability and Statistics (Fall 2018-2019).
• Intellingence for Cyber-Physical Systems (Spring 2017-2018).