Joint Astrophysical Colloquium
Machine Learning in Science: Opportunities, Challenges, and New Frontiers
Umberto Michelucci (Lucerne University of Applied Sciences and Arts)
Thursday 23/01/2025 @ 11:30, Sala Antonio Sollima (IV piano Battiferro)
Artificial intelligence is transforming scientific discovery, offering tools to analyze complex data, simulate phenomena, and automate tasks. In this talk, we will explore how machine learning applications in science differ fundamentally from classical areas such as the ones not strictly science related (face recognition for example). We will discuss its unique challenges, including interpretability, robustness, and handling limited data, alongside its immense potential to uncover insights in physics and beyond. By contrasting classical and scientific machine learning, this talk aims to equip scientists with a clear understanding of machine learning's limitations and opportunities in science, inspiring innovative scientific research.