Curriculum Vitae


You can download a PDF version of my CV here.

Jump to section: About Me | Education | Work Experience | Awards & Grants | Publications | Patents | Talks/Presentations/Posters | Student Supervision | Service to the Profession | Skills


About Me

I am a Senior Deep Learning Researcher at MediaTek Research in Cambridge/London (UK). I hold a PhD in the areas of Deep Learning and Artificial Intelligence, with a focus on Graph Representation Learning and Graph Neural Networks.

During my PhD I have been a Research Scientist Intern at Meta AI under the supervision of Dr. Vassilis Plachouras, and at Samsung AI Research under the supervision of Dr. Efthymia Tsamoura.

I've also been a visiting researcher in Professor Pietro Liò's group at the University of Cambridge, a visiting research in Dr. Bastian Rieck's group at Helmholtz Munich. Before starting my PhD I gained some industry experience with a six months internship developing Deep Learning algorithms for digital pathology, and spending 2 months as a Machine Learning consultant.

Education

Ph.D. in Information Engineering

University of Padova
My research was in the area of Deep Learning, with a particular focus on Graph Neural Networks and Graph Representation Learning.

Thesis: "Improving the Effectiveness of Graph Neural Networks in Practical Scenarios" [Full Text]
Supervisor: Prof. Fabio Vandin

October 2019 - March 2023

Master's Degree in Computer Engineering · 110/110 with honors

University of Padova
I focused my studies in the areas of Machine Learning, Data Mining, and Algorithmics.

Thesis: "A Deep Learning Model for Personalised Human Activity Recognition." [Full Text]
Advisor: Prof. Fabio Vandin

September 2016 - February 2019

Bachelor's Degree in Information Engineering

University of Padova

Thesis: "Algorithms for the determination of node centralities in a graph."
Advisor: Prof. Pietracaprina Andrea Alberto

September 2013 - June 2016

Work Experience

Senior Deep Learning Researcher

MediaTek Research · Cambridge/London, United Kingdom

Working in the AI research arm of MediaTek on both applied and fundamental research projects.

May 2023 - Present

Research Scientist Intern

Meta AI · London, United Kingdom

Worked on leveraging multi-modal data for tackling the cold-start problem in recommendation systems. A technical report regarding this project is available on arXiv.

September 2022 - December 2022

Visiting Researcher

Helmholtz Munich · Munich, Germany

As a recipient of the Helmholtz Visiting Research Grant, awarded by the Helmholtz Information and Data Science Academy, I have performed research at the intersection of Graph Neural Networks and Topological Data Analysis.
Supervisor: Dr. Bastian Rieck.

August 2022

Visiting Researcher

University of Cambridge · Cambridge, United Kingdom

Continued my research on Graph Neural Networks. In more detail I have worked on the problem of size-generalization, and on topological techniques to capture higher-order structures in graphs.
Supervisor: Professor Pietro Liò.

April 2022 - July 2022

Research Intern

Samsung AI Research · Cambridge, United Kingdom

I worked on neurosymbolic approaches combining Deep Learning and Logical Reasoning. In more detail, the research done at Samsung focused on the development of a logic-based loss function for deep learning models with the goal of injecting commonsense knowledge into scene graph generation models. This work has led to a publication (at AAAI 2023) and a patent.
Supervisors: Dr. Efthymia Tsamoura.

January 2021 - July 2021

Research Fellow

University of Padova · Padova, Italy

Project: "Machine Learning for Temporal Data".
Supervisor: Professor Fabio Vandin.
The research project focused on the development of novel Deep Learning frameworks for multimodal times series and has led to a publication on the IEEE Sensors Journal.

Aril 2019 - September 2019

Data Scientist, Machine Learning Engineer

Machine Learning Reply · Milan, Italy

During my time at Machine Learning Reply I had the chance to work for important clients on machine learning related projects. In particular I contributed to:

  • the development of a chatbot that aids financial traders in their operations.
  • the development of an automatic system for the analysis of documents and invoices.
The main technologies involved were: Python, Rasa, Google Cloud Vision, Java.

January 2019 - February 2019

Data Science, Deep Learning Algorithm Development Intern

Philips Digital & Computational Pathology · Belfast, United Kingdom

I worked in the team responsible for the development of algorithms that aid pathologists in the analysis of medical slides inside the TissueMark application. This implied the creation, training and validation of Deep Learning models and the engineering, processing and analysis of data (which was mainly composed of medical slides and relative metadata).
In more detail, my tasks included:

  • The creation of an automatic system for the tuning of hyperparameters for Deep Learning models.
  • The development of the algorithm for the identification of the appropriate tumour regions for macrodissection for lung tissue slides (which included the analysis and engineering of data and the training and validation of Deep Learning models).
  • The development of technical tasks to improve the company's Deep Learning framework infrastructure.
The main technologies involved were: Python, Keras, TensorFlow.

July 2018 - December 2018

Awards & Grants

  • Helmholtz Visiting Researcher Grant - 08/2022
    Grant awarded by the Helmholtz Information and Data Science Academy (HIDA), as part of the Helmholtz Association, to support a research stay at a Helmholtz centre.
  • Fondazione Luciano Iglesias Scholarship - 07/08/2020
    Award given to the 10 best M.Sc. graduates in Computer Engineering at the University of Padova in 2019.
  • Full PhD Scholarship from the Department of Information Engineering (University of Padova) - 2019-2022
  • ERASMUS+ Traineeship Grant - July 2018-December 2018

Publications

  • Exact, Tractable Gauss-Newton Optimization in Deep Reversible Architectures Reveal Poor Generalization
    Davide Buffelli*, Jamie McGowan*, Wangkun Xu, Alexandru Cioba, Da-shan Shiu, Guillaume Hennequin, Alberto Bernacchia, Thirty-Eighth Annual Conference on Neural Information Processing Systems (NeurIPS), 2024.
    [Paper coming soon]
  • Deep Equilibrium Algorithmic Reasoning
    Dobrik Georgiev, JJ Wilson, Davide Buffelli, Pietro Liò, Thirty-Eighth Annual Conference on Neural Information Processing Systems (NeurIPS), 2024.
    [Paper coming soon]
  • CliquePH: Higher-Order Information for Graph Neural Networks through Persistent Homology on Clique Graphs
    Davide Buffelli*, Farzin Soleymani*, Bastian Rieck, Preprint, 2024.
    [PDF (arXiv)]
  • The Deep Equilibrium Algorithmic Reasoner
    Dobrik Georgiev, Pietro Liò, Davide Buffelli, CVPR Workshop on Multimodal Algorithmic Reasoning, 2024. (Spotlight)
    [PDF (arXiv)]
  • Is Meta-Learning the Right Approach for the Cold-Start Problem in Recommender Systems?
    Davide Buffelli, Ashish Gupta, Agnieszka Strzalka, Vassilis Plachouras, Preprint, 2023.
    [PDF (arXiv)]
  • Improving the Effectiveness of Graph Neural Networks in Practical Scenarios
    Davide Buffelli, PhD Thesis, University of Padova, 2023.
    [Full Text]
  • Scalable Theory-Driven Regularization of Scene Graph Generation Models
    Davide Buffelli*, Efthymia Tsamoura*, Thirty-Seventh AAAI Conference on Artificial Intelligence (AAAI), 2023.
    [Paper] [PDF (arXiv)]
  • SizeShiftReg: a Regularization Method for Improving Size-Generalization in Graph Neural Networks
    Davide Buffelli, Pietro Liò, Fabio Vandin, Thirty-sixth Conference on Neural Information Processing Systems (NeurIPS), 2022.
    [Paper] [PDF (arXiv)] [Code]
  • Graph Representation Learning for Multi-Task Settings: a Meta-Learning Approach
    Davide Buffelli, Fabio Vandin, International Joint Conference on Neural Networks (IJCNN), 2022. (Oral)
    [Paper] [PDF (arXiv)] [Code]
  • The Impact of Global Structural Information in Graph Neural Networks Applications
    Davide Buffelli, Fabio Vandin, Data (special issue "Knowledge Extraction from Data Using Machine Learning"), 2022.
    [Paper] [PDF (arXiv)] [Code]
  • Extending Logic Explained Networks to Text Classification
    Rishabh Jain, Gabriele Ciravegna, Pietro Barbiero, Francesco Giannini, Davide Buffelli, Pietro Liò, The 2022 Conference on Empirical Methods in Natural Language Processing (EMNLP), 2022.
    [Paper] [PDF (arXiv)]
  • Attention-Based Deep Learning Framework for Human Activity Recognition with User Adaptation
    Davide Buffelli, Fabio Vandin, IEEE Sensors Journal, 2021.
    [Paper] [PDF (arXiv)] [Code]
  • A Meta-Learning Approach for Graph Representation Learning in Multi-Task Settings
    Davide Buffelli, Fabio Vandin, NeurIPS Workshop on Meta-Learning (MetaLearn), 2020.
    [PDF] [PDF (arXiv; with Appendix)] [Video] [Slides] [Poster] [Code]

Patents

  • Method and System for Scene Graph Generation
    Davide Buffelli, Efthymia Tsamoura, (patent pending), 2022.

Talks/Presentations/Posters

  • Invited Talk - "The Problem of Size-Generalization in Graph Neural Networks"
    Presented at the Artificial Intelligence Research Group Talks (Computer Laboratory) (University of Cambridge).
    04/07/2022 - [Link] [Video]
  • Invited Talk - "Word Embeddings & Graph Neural Networks for Automatic Reasoning over Knowledge Graphs"
    Presented at the Word Embedding Reading Group (University of Padova).
    25/05/2020 - [Slides] [Video]
  • Poster (Refereed Workshop) - "Are Graph Convolutional Networks Fully Exploiting Graph Structure?"
    Presented at the ELLIS Workshop on Geometric and Relational Deep Learning.
    24/02/2020 - [Slides] [Poster Video]

Student Supervision

  • Master's Thesis Supervision: Matteo Terranova ("Study of Regularization Techniques for Semi-Supervised Learning on Graphs with Graph Convolutional Networks"; co-supervised with Prof. Fabio Vandin, 2020).

Service to the Profession

Reviewer Conferences: RECOMB 2020, ISMB 2020, KDD 2020, ICDM 2020, NeurIPS Workshop on Meta-Learning (MetaLearn) 2020-2021, TheWebConf 2021, NeurIPS I Can't Believe It's Not Better! (ICBINB) Workshop 2021.
Journals: ACM Transactions on Information Systems, IEEE Sensors Journal.

Skills

Programming Languages & Tools

Proficient: Python.
Familiar: Java, C, Apache Spark, SQL, PostgreSQL, Objective-c, Swift.
I worked extensively, both in an academic and in a professional environment, with the main Machine Learning and Deep Learning libraries such as TensorFlow, PyTorch, Keras, Pandas, scikit-learn.

Workflow
  • Agile Development & Scrum