A fully funded 3-year Ph.D. position in Causal AI within the National PhD AI Program at the University of Pisa (Italy) is sponsored by the Ubiquitous Internet (UI) research group of IIT-CNR. We are looking for a highly motivated Ph.D. candidate with a strong academic background to join our research team and work on this Ph.D. topic under our supervision.
This scouting notice aims to raise awareness of the upcoming Ph.D. admission call that will be officially opened by the University of Pisa in June. We are currently collecting expressions of interest.
Research Topic
Traditional machine learning approaches primarily focus on correlation-based learning, identifying statistical associations between variables. Shifting from correlations to causal relationships is one of the most promising directions toward AI that is more robust, interpretable, and useful in practice.
This research explores how heterogeneous devices (smartphones, wearables, IoT sensors, and edge systems) can be used as a substrate for causal learning in the wild. The idea is to leverage data naturally collected from distributed, device-rich environments (smartphones, wearables, and IoT sensors) as a substrate for causal learning frameworks that can extract causal knowledge from observational data in real-world settings .
Research Focus & Methodology
The target applications focus on pervasive systems, where the heterogeneity and scale of device-generated data create both unique challenges and opportunities for causal reasoning. In these settings, causal knowledge can play a key role in improving decision-making and adaptive behavior — for instance, enabling systems to generalize across different devices and environments, act robustly under uncertainty, or understand the consequences of their actions rather than merely reacting to observed patterns.
Depending on the background and interests of the candidate, research activities may include:
- Theoretical modeling of causal inference in (distributed) AI settings.
- Algorithm and system design for deploying causal learning on pervasive devices.
- Using causal representations to improve decision-making, planning, or generalization under uncertainty.
- Experimental evaluation through simulations and real-world deployments.
Ideal Candidate Profile:
* MSc in Computer Science, Mathematics, Physics, or a related field
* Strong foundation in probability, statistics, and machine learning
* Background or interest in causal inference, sequential decision-making, or pervasive systems
* Comfortable with Python and relevant frameworks
Who Should Apply?
This is a PhD position. The specific research direction adaptable to the candidate’s expertise.
If this sounds like something you'd enjoy working on, feel free to get in touch.