|
Calendar of Physics Talks Vienna
| Machine learning for gravitational-wave inference at scale |
| Speaker: | Ippocratis Saltas (Czech Academy of Sciences) |
| Abstract: | Gravitational-wave astronomy promises a major advance in fundamental physics, but it also poses severe computational challenges. Signals are weak and buried in noise, detector noise is non-stationary, and standard Bayesian inference can require days to weeks per event. With next-generation detectors, inference must scale to longer signals, higher event rates, and real-time analysis, making traditional approaches increasingly prohibitive. In this talk, I will briefly introduce the basics of gravitational waves and outline how machine learning can play an enabling role by dramatically reducing computational cost. I will then review key challenges and discuss recent progress toward models with transfer-learning capabilities that remain reliable across detectors and waveform modelling, with relevance to emerging foundation-model efforts in physics. |
| Date: | Thu, 12.02.2026 |
| Time: | 14:15 |
| Duration: | 60 min |
| Location: | Seminarraum AC 02 - 2, Hauptgebäude (Karlsplatz 13), Stiege 5, 2. Stock, https://maps.tuwien.ac.at/?q=AC0240 |
| Contact: | Andreas Ipp, Ankit Aggarwal |
|