Solving high-dimensional parameter inference: marginal posterior densities & Moment Networks
From a 100-dimensional parameter space, directly estimate the mean and covariance for pairs of parameters with Moment Networks:
MomentNetwork_demo/moment_network_100D.ipynb
Google Colab notebook can be run in browser with GPU acceleration
(modules loaded with !pip install 'git+https://github.com/NiallJeffrey/MomentNetworks.git'
)
Demonstrate marginal posterior moment estimation using high-dimensional LIGO-like gravitational wave time series