Searching for bias and correlations in a Bayesian way

TL;DR

This paper introduces a Bayesian approach using internal robustness to detect biases and hidden correlations in large cosmological datasets, enhancing measurement accuracy in a model-independent manner.

Contribution

It extends previous methods by applying internal robustness to identify biased data subsets and correlations without relying on specific models.

Findings

Effective detection of biased data subsets

Identification of hidden correlations in datasets

Improved accuracy in cosmological parameter inference

Abstract

A range of Bayesian tools has become widely used in cosmological data treatment and parameter inference (see Kunz, Bassett & Hlozek (2007), Trotta (2008), Amendola, Marra & Quartin (2013)). With increasingly big datasets and higher precision, tools that enable us to further enhance the accuracy of our measurements gain importance. Here we present an approach based on internal robustness, introduced in Amendola, Marra & Quartin (2013) and adopted in Heneka, Marra & Amendola (2014), to identify biased subsets of data and hidden correlation in a model independent way.