Simone is co-head of the Applied AI Research Lab at the Lucerne University of Applied Sciences and Arts. His main interest is the application of AI to health and medical technologies. Simone completed a PhD in Theoretical Particle Physics with focus on numerical and analytic computational methods at ETH Zürich. His education and professional stations include the University of Milano, Stanford, DESY, and Durham. Since 2019, Simone is working at the boundary between industry and academia, with the goal to translate the most recent advances in Machine Learning into technologies that improve life quality. He supports R&D developers, clinicians, and decision-makers within organisations from start-ups to large corporations in designing AI products, data acquisition, model training and auditing. He regularly lectures and speaks at conferences both for specialists and the general public.
Real-world data collections almost inevitably contain issues which decrease their value, for example by deteriorating performance or undermining trust.
Whether data are gathered via user inputs, sensors, or careful clinical protocols, they will have artefacts such as near-duplicate, off-topic, or wrongly labelled samples.
Thankfully, there are ways to detect inconsistencies using modern techniques like self-supervision.
This enables to find issues during collection, before training, or even during deployment of AI models on data streams of increasing size, and involving human labour only as much as needed.
For conference production and speaking opportunites:
Für die Produktion von Konferenzen und die Möglichkeit, Vorträge zu halten:
For sponsorship and exhibition opportunities: