[TUTORIAL] Amortized LFI of cluster masses with TensorFlow Probability #6
Description
Amortized LFI of galaxy cluster masses with TensorFlow Probability
Learn to use TensorFlow Probability to get posterior mass estimates of cluster masses
using dynamical mass measurements.
Contacts: @EiffL
Participants:
Detailed description
This tutorial illustrates how to build an amortized inference model for estimating
galaxy cluster masses using TensorFlow Probability (TFP). It covers the definition of a conditional density estimator using Keras, understanding the impact of the proposal distribution used to build the training dataset, and learning how to get proper Bayesian posteriors.
Ressources: colab notebook