We kindly invite you to the seminar Tree Based Search for Self-Driving Virtual Labs by Stefan Bauer.
The seminar will be held on 17th April 2024 at 12.00 pm at the IDEAS NCBR Conference Room.
Abstract:
A tremendous range of design tasks in materials, physics, and biology can be formulated as finding the optimum of an objective function. We present a tree based search method for derivative-free optimization that enables accelerated optimal design of high-dimensional complex systems. Our method demonstrates applicability to a wide range of real-world complex systems spanning materials, physics, and biology, considerably outperforming state-of-the-art algorithms. This enables efficient autonomous knowledge discovery and opens the door towards scalable self-driving virtual laboratories.
Bio:
Stefan Bauer is currently an associate professor at TU Munich and senior PI at Helmholtz AI. The research group is one of the few worldwide, which is successfully working on the intersection of causality and deep learning not only in simulation but in concrete real-world downstream tasks. The work is regularly published at top machine learning conferences NeurIPS, ICML (best paper award), ICLR as well as in PNAS or Science and recognized through a CIFAR Azrieli Global Scholar Fellowship and significant international media coverage e.g. in MIT Technology Review, The Royal Chemical Society, El País or IEEE Spectrum.
***
We encourage you to participate in the event in person. Participation via the Zoom platform is also possible. Please send an e-mail to registration@ideas-ncbr.pl with your name, surname and affiliation.
https://us06web.zoom.us/j/88598770279?pwd=ddPQrhhCXuq9pyDkvoHMUqaW74oCZb.1
Meeting ID: 885 9877 0279
Passcode: 847914
