We are thrilled to welcome Dr Eleonora Giunchiglia to the I-X Team!

 

Eleonora joined Imperial in June 2024 as a lecturer at I-X and the Department of Electrical and Electronic Engineering. We recently had the opportunity to ask her a few questions about her research and her non-academic interests.

Welcome to the I-X team! Could you tell us a bit about yourself and your academic background?

Thanks for the warm welcome! I am joining I-X and the Department of Electrical and Electronic Engineering as a Lecturer. Previously, I completed my PhD at the University of Oxford in 2022, followed by a post-doctoral research position at TU Wien. My expertise lies at the intersection of machine learning and logic, enabling me to develop new methods by integrating techniques from both fields.

What are your main research interests and what do you work on currently? 

Since the beginning of my PhD, my goal has been to build deep learning-based systems that are safe and trustworthy. In my research, this translates to developing deep learning models that are compliant by design with a set of formally defined requirements, ensuring the desired behaviour of the model. While this concept has numerous applications, I have primarily focused on autonomous driving and synthetic data generation.

What motivated you to work in this area?

Honestly, my motivation stemmed from frustration with deep learning, which, despite its success in many cases, can occasionally make unexpected mistakes. During my MSc thesis, I developed a model predicting survival curves for ICU patients. These curves must monotonically decrease over time, meaning that if a model predicts that the patient will survive till next month with a probability equal to 0.5, it then cannot predict a 0.8 survival probability until next year. Despite high accuracy and performance metrics, the model I developed sometimes violated this very simple requirement, making nonsensical predictions. I was unable to instruct the model to adhere to this requirement, thus I decided to pursue a PhD focused on solving this problem on a larger scale.

What attracted you to working at I-X?

Given my research, I-X represents the perfect place to be. I-X is a unique institution that attracts some of the best talent globally and hosts academics from diverse backgrounds. Each of these academics is exploring AI from different angles and applying it to various settings beyond my area of expertise. My hope is to collaborate with them to unlock the full potential of developing deep learning models that are compliant by design with specified requirements.

What are your interests outside of work? Could you tell us an interesting fact about yourself?

My favourite activities include cooking and enjoying meals with family and friends, especially outdoors and under the sun. Ideally, these meals are followed by a good book or an interesting movie, making my day perfect. A fun fact about me is that, as a child, I really got into reading, to the point that my parents prohibited me from buying new books. This led me to discover the pleasure of re-reading my favourite ones over and over again.