For UvA Master students: you can drop me an e-mail if you are looking for thesis projects on machine learning/deep learning.

I am a postdoctoral researcher with Max Welling at the UvA-Bosch DELTA Lab (Deep Learning Technologies Amsterdam), a joint effort of the University of Amsterdam and Bosch, where we work on fundamental research in deep learning for computer vision and automotive.

Previously, I did research and engineering on distributed machine learning and privacy at N1analytics, part of Data61 CSIRO. I completed my PhD in machine learning at the Australian National University with NICTA, advised by Richard Nock and Tiberio Caetano. My thesis was about weakly supervised learning, e.g. labels corrupted by noise or by high-level aggregation, touching on learning theory, privacy and deep learning. I received my BSc and MSc from Polytechnic University of Milan with a thesis on Nash equilibrium computation with Nicola Gatti.

I like startups. I co-founded and advice Waynaut, among the few companies turning up the heat (Forbes) of the Italian startup environment.

News

2017

• Sep - I have published a new post on recent insights for harvesting signals from unsupervised data
• Sep - I have moved to UvA-Bosch DELTA Lab (Deep Learning Technologies Amsterdam) as a postdoctoral researcher
• Aug - We have published a follow up blog post on how to use encrypted models for prediction
• Aug - Slides of our ICML17 workshop talk on entity resolution and logistic regression with encrypted data are online
• Jul - We published a first blog post on federated machine learning and Paillier homomorphic encryption
• Jun - ICML17 paper accepted at workshop on Private and secure machine learning: Privacy-preserving entity resolution and logistic regression on encrypted data
• Jun - My PhD thesis is now online
• May - I am a recipient of the Young Researcher Award sponsored by CVPR17
• Feb - CVPR17 paper accepted as oral: Making Deep Neural Networks Robust to Label Noise: a Loss Correction Approach