Explaining the world, one model at a time.

I'm Mattia Setzu, a junior researcher at the Department of Computer Science of the University of Pisa, Italy. I focus on Explainable AI, with the broad research goal of automating the understanding of complex machine learning models.

Talks

A list of my talks and presentations.

Seminar on Transformers at Scuola Normale

15/11/2022

GLocalX @Intesa

1/12/2022

Invited talk at Banca Intesa.

Bad researcher, bad developer

15/11/2022

Seminar on coding for research.

Reading fast and slow

8/11/2022

Seminar on literature management for research.

Ph.D. defense

Seminar on Transformers at Scuola Normale

Joint presentation with Carlo Metta.

GLocalX at Imperial

17/3/2022

TriplEx at CogMI2021

14/12/2021

Ph.D. discussion

GLocalX

GLocalX, from “GLocalX - From Local to Global Explanations of Black Box AI Models”, presented at IJCAI 2021.

10/8/2021

Ph.D. proposal

10/11/2021

Understanding the role of individual units in a deep neural network

KDD Lab seminars on Explainability.

Presentations about the ECML/PKDD 2019 conference

5/11/2019

Presentation on several papers from the ECML/PKDD 2019 conference.

  • DeepProbLog: Neural Probabilistic Logic Programming
  • Agnostic Feature selection
  • Fast Gradient Boosting Decision Trees with Bit-Level Data Structures

Agnostic Feature Selection

5/11/2019

Presentation on Agnostic Feature Selection, a Best Student Paper Award by Guillaume Doquet and Michèle Sebag. Published at ECML/PKDD 2019.

PhD Thesis Proposal

22/10/2019

PhD Thesis Proposal on Explainability, titled “Opening the Black Box: Empowering Machine Learning Models with Explanations”.

Global Explanations with Local Scoring

20/10/2019

Poster for “Global Explanations with Local Scoring”, my publication at XKDD, an ECLM/PKDD workshop on Explainability.

Transfair Fairness

3/7/2019

Presentation slides on myself, my research group and a weird research idea.

5min Local to Global

11/6/2019

5 minutes presentation to introduce the incoming director of ISTI-CNR to my research interests.

Tackling Explainability in a Local to Global Setting

7/5/2019

Presentation in the “Mauriana Pesaresi” cycle of seminars.

Master’s Thesis

5/10/2018

Defence of my master’s thesis.

Spanner, Google’s distributed database

Presentation on Spanner, a distributed database developed by Google. The presentation was part of the exam ‘Parallel and Distributed algorithms’ at UniPI.

Codeontology, A Java Semantic Database

Seminary at UniCA in which I presented preliminary results of my bachelor’s thesis.