Three USI projects win the prestigious SNSF Ambition competition

Three projects from Università della Svizzera italiana (USI), one affiliated to the Ente Ospedaliero Cantonale (EOC), have won the prestigious SNSF Ambizione call. Directing the winning projects are Emanuele Guidotti, Post-doc at the Faculty of Economics; Alberto Martín López, Post-doc at the Faculty of Informatics; and Anna Castelnovo, Private Lecturer at the Faculty of Biomedical Sciences.

Emanuele Guidotti's project, conducted within the Institute of Finance (IFin) of the Faculty of Economics, is titled "Financial Market Microstructure, Big Data, and AI". The idea for the project stems from the consideration that, to date, little is known about how to estimate liquidity in the past when high-frequency data are not available and how to simulate the impact of AI on financial markets in the future. At the root of both problems is a lack of practical understanding of how an asset's price is determined in financial markets, which prevents connecting high-frequency data to low-frequency data and hinders realistic simulations. For the past two years, Emanuele Guidotti has been working on a price formation model that forecasts and explains new verifiable relationships between key variables in financial markets and that he hypothesises can be used to link the microstructure of financial markets to macroscopic effects. Therefore, the project aims to finalise and use the model to improve forecasting and simulation in critical applications. The three-part proposal involves empirical analyses using the NYSE Trade and Quote (TAQ) database. The project involves collaboration with the world's leading experts in market microstructure and artificial intelligence: Prof. Albert S. (Pete) Kyle (Robert H. Smith School of Business, University of Maryland); Prof. Albert Menkveld (Vrije Universiteit Amsterdam); Prof. Björn Hagströmer (Stockholm Business School); Prof. Stefano M. Iacus (Harvard University); Prof. Nicolò Cesa-Bianchi (University of Milan); and Prof. Semyon Malamud (École Polytechnique Fédérale de Lausanne, EPFL).

The second winning project, led by Alberto Martín López within the Software Institute (SI) of the Faculty of Informatics, is titled "SAPIENS: Supporting the API Lifecycle with Neuro-Symbolic AI". As software becomes woven into the very fabric of our world, web application programming interfaces (APIs) become the dots, the glue that holds everything together; they enable heterogeneous systems to communicate over the Internet and provide access to all kinds of data and services. Today, artificial intelligence (AI) is a key ingredient of software development, with tools such as ChatGPT and GitHub Copilot changing the way software is created. However, this also poses challenges, as AI-driven development falls short in the field of web APIs. To address the challenge of simplifying a wide variety of complex tasks in the API lifecycle, Alberto Martín López hypothesises that neuro-symbolic AI (NSAI) is the way forward. NSAI combines the power of neural networks, such as pre-trained code models and large language models (LLM), with the robustness and determinism of symbolic techniques, such as invariant detection and constraint programming. The project's goal, which is to be achieved through the investigation of multiple research questions, is twofold: to empower developers in developing, maintaining and consuming web APIs and to improve the reliability of real-world services.

The last project, affiliated with the EOC, is led by Anna Castelnovo and is titled "Mental and Neural Correlates of NREM Parasomnias in Children". NREM parasomnias, commonly referred to as "waking disorders," include conditions such as somnambulism (sleepwalking), night terrors, and confusional awakenings. These disorders involve recurrent and unpredictable episodes of partial awakening from deep sleep, during which individuals experience a coexistence of sleep and wakefulness, leading to altered states of consciousness. While these events are much more common in children, most studies conducted to date have focused on adults. Moreover, diagnosis is sometimes complex, especially due to the difficulties of differential diagnosis with some forms of epilepsy. Currently, there are no reliable tools to predict or prevent these episodes in children, which can have a major impact on their well-being and that of their families. Available treatment options are limited and based on knowledge that is still fragmentary.This project aims to address this gap by systematically studying the changes that occur in children's brains before, during, and after an episode. It will utilise high-density electroencephalograms (EEG) and engage in conversations with children immediately following each episode to identify any mental experiences, such as dream images or hallucinatory states. The project collaborates with IDSIA USI-SUPSI and ZHAW to analyse the data using machine learning, with the University of Zurich to develop the neurostimulatory aspects, and with two Italian academic centres for the video analysis of the episodes.

The SNSF Ambizione call for proposals is aimed at young researchers (within a maximum of four years after obtaining their doctorate) and allows them to carry out and manage an independent research project at a Swiss institution. The funding covers the researcher's salary and up to a maximum of CHF 400,000 for research-related costs. The call is extremely competitive, with a success rate of around 18%.