Invited

Blackboard talk done for the 2023 “Building a physical understanding of galaxy evolution with data-driven astronomy” KITP workshop https://datadrivengalaxyevolution.github.io/.

Done for the 2023 “Building a physical understanding of galaxy evolution with data-driven astronomy” KITP workshop https://datadrivengalaxyevolution.github.io/. Recording can be found by following this link.

Leading the Accelerated Forward Modelling discussion on our models for different aspects of the dark matter - galaxy connection

I will introduce a new Machine Learning based model for mapping between dark matter and galaxies, called Mangrove. I’ll show improvements in both precision and speed compared to other similar methods, and discuss how Mangrove opens up new avenues for efficiently learning from simulations.

For the first time ever, we show that it is possible to use full merger histories to emulate SAMs with Machine Learning and vastly outperform the state-of-the-art in the field.

For the first time ever, we show that it is possible to use full merger histories to emulate SAMs with Machine Learning.

Doing fast and precise forward simulations of galaxies will be of fundamental importance to fulfilling the LtU collaboration goals. In this talk I’ll show how we can both get faster, better galaxies, and understand them better with a GNN-based emulator

For the first time ever, we show that it is possible to use full merger histories to emulate SAMs with Machine Learning.