Research Overview

Research at the Intersection of Astrophysics and Machine Learning

Research Themes

The institute's main research activities are focused on developing rigorous Bayesian inference frameworks that employ deep learning to accelerate parameter estimation and increase the accuracy of the inferred posteriors.

Black Holes

Breakthrough discoveries about the most fascinating objects in the Universe, black holes, ultimately revealing how these objects act as the beating heart of all galaxies.

Lensing for Dark Matter

Discovering the particle properties of dark matter and probing cosmology with strong gravitational lensing and machine learning.

Cosmology

Revealing the origin of the Universe, the nature of its components and how they affect its evolution, and its eventual fate.

Inference

Using neural networks to create new statistical analyses for complex, high-dimensional astrophysical datasets.

Publications

Featured Projects

Learning the Universe

Research Areas : Cosmological parameters, dark matter, dark energy, large sky surveys, simulation-based inference

Future Lens

Research Areas : Strong gravitational lensing, dark matter, Hubble constant, dark energy, computer vision, AGN

Gaia Data

Research Area : Anomaly detection, unsupervised learning, density estimation, point clouds, HR diagram, Milky Way stream

Turbulence

Research Area : Subgrid physics, closure, Navier–Stokes equations, deep learning, recurrent models, Dedalus

X-Ray Astronomy

Research Area : Active Galactic Nuclei, intra cluster medium, X-ray binaries, galaxy clusters, deep learning