Mattia Cerrato

Office 03-621

Staudingerweg 9

55128 Mainz, Germany

I work as a Post-Doc in the Data Mining group at Johannes Gutenberg-Universität Mainz, under the guidance of Prof. Stefan Kramer. I am an Advisor/Tutor for the TOP-ML project, which deals with understanding the trade-offs between properties of machine learning algorithms beyond performance. Before that, I was a PhD. student in the Machine Learning Group at University of Torino, my hometown. Prof. Roberto Esposito was my advisor for three years.

My research interests are centered around deep neural networks - I mostly research algorithms to constrain and understand them better. I have published papers about algorithmic fairness, interpretability and connections between them. I care deeply about making sure that AI applications can be beneficial to everyone - or at least not harm anybody.

If you are an undergrad and are interested in working on deep learning, interpretability and fairness please check out the projects page.

news

Nov 20, 2022	My paper “Invariant Representations with Stochastically Quantized Neural Networks” has been accepted at AAAI23! This is work done alongside Marius Koppel, Roberto Esposito and Stefan Kramer. Read the preprint here.
Jun 23, 2022	Our preprint “Fair Interpretable Representation Learning with Correction Vectors” has been featured by the Montreal AI Ethics Institute. Read our summary here.
Mar 1, 2022	I am teaching the Machine Learning Tutorials during the Summer Semester 2022
Oct 10, 2021	I got my PhD. from the University of Torino. My thesis’ title is “Invariance in Neural Representations with applications to fairness and debiasing”.