Marco Mondelli wins €1 million Lopez-Loreta prize

The Lopez-Loreta prize, along with €1 million, is given annually to four recent master or PhD graduates from EPFL, ETH Zurich, the École Polytechnique in Palaiseau or the ISAE-SUPAERO in Toulouse to finance proposed projects over five years. The prize is funded by the Jean-Jacques and Felicia Lopez-Loreta Foundation for academic excellence.

Mondelli is currently a postdoctoral researcher in the Department of Electrical Engineering at Stanford University in California. His research interests include data science and machine learning, wireless communication systems, network information theory and modern coding theory. He received the Lopez-Loreta prize to support a five-year research project on the foundations of deep learning.

The goal of the project is to build theoretical foundations to guide the design of more scalable and efficient deep learning systems by addressing two complementary challenges: characterizing models that are algorithmically difficult to learn, and providing efficient algorithms with provable guarantees. He and his colleagues plan to validate their findings by conducting extensive numerical experiments on real-world datasets.

Mondelli completed his PhD in the EPFL School of Computer and Communication Sciences (IC) Communication Theory Laboratory (LTHC) led by Professor Rüdiger Urbanke. In 2017, he won an EPFL thesis distinction for his doctoral thesis on coding theory for wireless communications: “From Polar to Reed-Muller Codes Unified Scaling, Non-standard Channels, and a Proven Conjecture”.

In September, Mondelli will be joining the Institute of Science and Technology Austria (IST Austria) as an assistant professor.

“None of this would have been possible without the incredible support that I received both in terms of an incredibly stimulating research environment and in terms of accessibility to funding,” he said in a recent testimonial for the IC doctoral school, EDIC.