Advances in the Statistical Treatment of Systematic Errors across the Quantitative Sciences

In diverse fields of the quantitative sciences, including astronomy, physics, statistics, mathematics and data science, there is a well-established classification of data uncertainties into either random (statistical) errors, which stem from data variance, or systematic errors, which result from consistent biases affecting measurements in specific directions. This classical categorization is mirrored in machine learning with terms like "aleatoric" for random and "epistemic" for systematic uncertainties. Despite the recognition of these error types, there remains a significant need for developed, statistically sound methodologies to properly incorporate them into data analysis tasks including regression, goodness of fit tests, hypothesis testing, and parameter estimation.

This Research Topic aims to provide a comprehensive review of methods for integrating systematic errors across a wide range of scientific disciplines. Additionally, it seeks to identify and explore new theoretical and practical avenues for applying these errors to real-world data analysis challenges. The goal is dual: to augment the tools available to scientists and researchers for handling systematic errors and to spur further innovations in this area.

To further understand the nuances of systematic error handling within varied scientific domains, we encourage contributions addressing, but not limited to, the following specific themes:

• Development and evaluation of statistical methods for systematic errors in astronomy and the physical sciences, statistics, mathematics, and computer science
• Theoretical development of statistical methods for systematic errors
• Techniques and challenges in modeling systematic errors within biological and medical statistics
• Implications of systematic errors in economic and business data analysis, with a focus on econometrics
• Innovations in algorithmic approaches to handling systematic and epistemic errors in machine learning and artificial intelligence

We invite studies that are theoretical, applied, or a combination of both, aiming at solidifying a quantitative and statistically sound approach towards using and modeling systematic errors in various scientific practices.

Keywords: Astrostatistics, Data Science, Systematic Errors, Machine Learning, Artificial Intelligence, Epistemic Errors, Large Data, Regression, Hypothesis Testing, Biostatistics

Important Note: All contributions to this Research Topic must be within the scope of the section and journal to which they are submitted, as defined in their mission statements. Frontiers reserves the right to guide an out-of-scope manuscript to a more suitable section or journal at any stage of peer review.