arXiv subjects

David Salgado

Publications and source records attributed to David Salgado.

5 recordsLinked to original sources

arXiv2025

Streamlining business functions in official statistical production with Machine Learning

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).

stat.AP↗

arXiv2024

Leveraging Machine Learning for Official Statistics: A Statistical Manifesto

It is important for official statistics production to apply ML with statistical rigor, as it presents both opportunities and challenges. Although machine learning has enjoyed rapid technological advances in recent years, its application does not possess the methodological robustness necessary to produce high quality statistical results. In order to account for all sources of error in machine learning models, the Total Machine Learning Error (TMLE) is presented as a framework analogous to the Total Survey Error Model used in survey methodology. As a means of ensuring that ML models are both internally valid as well as externally valid, the TMLE model addresses issues such as representativeness and measurement errors. There are several case studies presented, illustrating the importance of applying more rigor to the application of machine learning in official statistics.

stat.ML↗

arXiv2021

A set of R packages to estimate population counts from mobile phone data

In this paper, we describe the software implementation of the methodological framework designed to incorporate mobile phone data into the current production chain of official statistics during the ESSnet Big Data II project. We present an overview of the architecture of the software stack, its components, the interfaces between them, and show how they can be used. Our software implementation consists in four R packages: destim for estimation of the spatial distribution of the mobile devices, deduplication for classification of the devices as being in 1:1 or 2:1 correspondence with its owner, aggregation for estimation of the number of individuals detected by the network starting from the geolocation probabilities and the duplicity probabilities and inference which combines the number of individuals provided by the previous package with other information like the population counts from an official register and the mobile operator penetration rates to provide an estimation of the target population counts.

cs.MS↗

arXiv2020

On new data sources for the production of official statistics

In the past years we have witnessed the rise of new data sources for the potential production of official statistics, which, by and large, can be classified as survey, administrative, and digital data. Apart from the differences in their generation and collection, we claim that their lack of statistical metadata, their economic value, and their lack of ownership by data holders pose several entangled challenges lurking the incorporation of new data into the routinely production of official statistics. We argue that every challenge must be duly overcome in the international community to bring new statistical products based on these sources. These challenges can be naturally classified into different entangled issues regarding access to data, statistical methodology, quality, information technologies, and management. We identify the most relevant to be necessarily tackled before new data sources can be definitively considered fully incorporated into the production of official statistics.

stat.OT↗

arXiv2006

Spin entanglement loss by local correlation transfer to the momentum

We show the decrease of spin-spin entanglement between two s=1/2 fermions or two photons due to local transfer of correlations from the spin to the momentum degree of freedom of one of the two particles. We explicitly show how this phenomenon operates in the case where one of the two fermions (photons) passes through a local homogeneous magnetic field (optically-active medium), losing its spin correlations with the other particle.

quant-ph↗