Hybrid Summary Statistics

We present a way to capture high-information posteriors from training sets that are sparsely sampled over the parameter space for robust simulation-based inference. In physical inference problems, we can often apply domain knowledge to define traditional summary statistics to capture some of the inf...

Full description

Saved in:
Bibliographic Details
Published in:arXiv.org 2024-10
Main Authors: T Lucas Makinen, Sui, Ce, Wandelt, Benjamin D, Porqueres, Natalia, Heavens, Alan
Format: Article
Language:English
Subjects:
Datasets
Inference
Information retrieval
Neural networks
Parameter robustness
Robustness (mathematics)
Statistics
Summaries
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We present a way to capture high-information posteriors from training sets that are sparsely sampled over the parameter space for robust simulation-based inference. In physical inference problems, we can often apply domain knowledge to define traditional summary statistics to capture some of the information in a dataset. We show that augmenting these statistics with neural network outputs to maximise the mutual information improves information extraction compared to neural summaries alone or their concatenation to existing summaries and makes inference robust in settings with low training data. We introduce 1) two loss formalisms to achieve this and 2) apply the technique to two different cosmological datasets to extract non-Gaussian parameter information.
ISSN:2331-8422