Streamlining business functions in official statistical production with Machine Learning

2025/10/28 by Barragán, Sandra, Pérez-Bote, Adrián, Sáez, Carlos +2
#Applications (stat.AP) #FOS: Computer and information sciences #Methodology (stat.ME)

paper · doi:10.48550/arxiv.2510.24394

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).

Citations

• Multilingual hierarchical classification of job advertisements for job vacancy statistics
• Cross-Validation: What Does It Estimate and How Well Does It Do It?
• Design-unbiased statistical learning in survey sampling
• RoBERTa: A Robustly Optimized BERT Pretraining Approach
• Bag of Tricks for Efficient Text Classification
• Total Survey Error: Past, Present, and Future

Related

• Machine Learning in Official Statistics
• Machine Learning for the Production of Official Statistics: Density Ratio Estimation using Biased Transaction Data for Japanese labor statistics
• Insights from Machine Learning for Evaluating Production Function Estimators on Manufacturing Survey Data
• Theory and Practice of Quality Assurance for Machine Learning Systems An Experiment Driven Approach
• Leveraging Machine Learning for Official Statistics: A Statistical Manifesto