FLORAH - Planting Better Merger Trees with Deep Generative Models - Tri Nguyen (MIT)

Explore the application of deep generative models to improve merger tree construction in galaxy formation studies through this 28-minute conference talk by Tri Nguyen from MIT. Delve into the FLORAH method, which uses machine learning to plant better merger trees, and understand its potential impact on astrophysical research. Learn about the background, objectives, and methodology of this innovative approach, including the use of Chat CBT simulations and training processes. Examine the results through power spectrum analysis and visualizations, and consider the future implications of this work for the field of galaxy evolution. Gain insights into how data-driven tools and astrostatistics are advancing our understanding of galaxy formation physics in the context of current and upcoming astronomical surveys.

Introduction
Background
How to plant merger trees
Pros and cons of 4D merger trees
Objectives
Chat CBT
Simulations
Training
Generation
Results
Plot
Simulation
Power Spectrum
Future Work
Summary
Discussion