Inaugural lecture: Machine Learning Theory in the Age of AI Revolution
Event details
| Date | 09.05.2023 |
| Hour | 17:15 › 18:00 |
| Speaker | Prof Florent Krzakala |
| Location | |
| Category | Inaugural lectures - Honorary Lecture |
| Event Language | English |
Abstract
As groundbreaking advancements in artificial intelligence emerge every week, the question arises: can theoreticians still contribute meaningfully to understanding these remarkable engineering feats? Traditional foundations, based on bias-variance tradeoffs, convex optimization, and data-agnostic theory, struggle to accommodate the complexities of modern methods. In this talk, we will explore how new ideas and directions at the intersection of high-dimensional probability, statistical physics, and information theory provide fresh perspectives on these challenges. By focusing on simple toy mathematically solvable models, we hope to gain a deeper understanding of the relationships between data, performance, and algorithm.
Bio
Florent Krzakala is a French physicist and applied mathematician, currently a professor at EPFL. His research focuses on using mathematical tools inspired by statistical physics to solve theoretical problems in physics, computer science, machine learning, statistics, and signal processing. He received his PhD in statistical physics jointly from Pierre and Marie Curie University and Paris-Sud University in 2002. After a postdoc with Giorgio Parisi in Rome, he was a lecturer at ESPCI in Paris and then a professor at Sorbonne University and the Ecole Normale Supérieure in Paris, where he also held the data science chair and founded the startup LightOn. He has held various visiting positions at universities in the United States and was named a junior fellow of the Institut Universitaire de France in 2015. Krzakala leads the Idephics laboratory at EPFL and is known in particular for his work on community detection, quantum annealing, and phase transitions in satisfiability and coloring problems, as well as high-dimensional statistics and algorithms.
As groundbreaking advancements in artificial intelligence emerge every week, the question arises: can theoreticians still contribute meaningfully to understanding these remarkable engineering feats? Traditional foundations, based on bias-variance tradeoffs, convex optimization, and data-agnostic theory, struggle to accommodate the complexities of modern methods. In this talk, we will explore how new ideas and directions at the intersection of high-dimensional probability, statistical physics, and information theory provide fresh perspectives on these challenges. By focusing on simple toy mathematically solvable models, we hope to gain a deeper understanding of the relationships between data, performance, and algorithm.
Bio
Florent Krzakala is a French physicist and applied mathematician, currently a professor at EPFL. His research focuses on using mathematical tools inspired by statistical physics to solve theoretical problems in physics, computer science, machine learning, statistics, and signal processing. He received his PhD in statistical physics jointly from Pierre and Marie Curie University and Paris-Sud University in 2002. After a postdoc with Giorgio Parisi in Rome, he was a lecturer at ESPCI in Paris and then a professor at Sorbonne University and the Ecole Normale Supérieure in Paris, where he also held the data science chair and founded the startup LightOn. He has held various visiting positions at universities in the United States and was named a junior fellow of the Institut Universitaire de France in 2015. Krzakala leads the Idephics laboratory at EPFL and is known in particular for his work on community detection, quantum annealing, and phase transitions in satisfiability and coloring problems, as well as high-dimensional statistics and algorithms.
Links
Practical information
- General public
- Registration required
Organizer
- Faculté STI – Décanat & Institut IEM