Streamlining business functions in official statistical production with Machine Learning
Applications
2025-10-29 v1 Methodology
Abstract
We provide a description of pilot and production experiences to streamline some business functions in the official statistical production process using statistical learning models. Our approach is quality-oriented searching for an improvement on accuracy, cost-efficiency, timeliness, granularity, response burden reduction, and frequency. Pilot experiences have been conducted with data from real surveys in Statistics Spain (INE).
Cite
@article{arxiv.2510.24394,
title = {Streamlining business functions in official statistical production with Machine Learning},
author = {Sandra Barragán and Adrián Pérez-Bote and Carlos Sáez and David Salgado and Luis Sanguiao-Sande},
journal= {arXiv preprint arXiv:2510.24394},
year = {2025}
}
Comments
42 pages, 14 figures, preprint version to appear as a chapter of F. Dumpert (ed.), Foundations and Advances of Machine Learning in Official Statistics