What does this research mean for the field?

Established probabilistic forecast reconciliation tools, such as covariance-weighted reconciliation and nonnegative post-processing, can be successfully operationalized into a transparent, fully reproducible statistical production pipeline for official tourism statistics. Novelty: ClaimNovelty.METHODOLOGICAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

What question did this study set out to answer?

Evaluate a fully reproducible probabilistic forecast reconciliation pipeline to ensure coherence and nonnegativity in official hierarchical tourism statistics.

September 5, 2026Statistical Journal of the IAOS

A reproducible reconciliation pipeline for probabilistic official tourism statistics: Evidence from Eurostat

View Full Paper

Ask AI

Bookmark

Authors

DKDo Yeon Kim

Discussion

Member takes

Overview

Empirical evaluation demonstrates coherent probabilistic forecast reconciliation across European tourism hierarchies, indicating reproducible pipelines improve official statistical accuracy.

Key Points

Evaluate a fully reproducible probabilistic forecast reconciliation pipeline to ensure coherence and nonnegativity in official hierarchical tourism statistics.
Analyzed Eurostat monthly tourism-nights series (tour_occ_nim) structured as a two-level hierarchy of 26 member states and their aggregated sum.
Generated rolling-origin probabilistic forecasts across horizons of 1, 3, 6, and 12 months using covariance-weighted reconciliation (MinT-W) and nonnegative post-processing (NonNeg).
Assessed forecast performance using the Continuous Ranked Probability Score (CRPS), Logarithmic Score (LogS), and Diebold–Mariano tests with Holm adjustment in a deterministic pipeline.
Operationalized established reconciliation estimators into a transparent, deterministic pipeline that guarantees mathematical coherence and nonnegativity across hierarchical series.
Demonstrated statistically rigorous probabilistic evaluation across multi-month horizons using proper scoring rules, fully regenerable via a standardized reproduction package.

Cite This Study

Do Yeon Kim (2026) studied this question.

synapsesocial.com/papers/6a9bd4046b95aff0620eb469 https://doi.org/10.1177/18747655261484249

View Full Paper

Ask AI

Bookmark

September 5, 2026Statistical Journal of the IAOS

A reproducible reconciliation pipeline for probabilistic official tourism statistics: Evidence from Eurostat

View Full Paper

Ask AI

Bookmark

Authors

DKDo Yeon Kim

Discussion

Member takes

Overview

Empirical evaluation demonstrates coherent probabilistic forecast reconciliation across European tourism hierarchies, indicating reproducible pipelines improve official statistical accuracy.

Key Points

Evaluate a fully reproducible probabilistic forecast reconciliation pipeline to ensure coherence and nonnegativity in official hierarchical tourism statistics.
Analyzed Eurostat monthly tourism-nights series (tour_occ_nim) structured as a two-level hierarchy of 26 member states and their aggregated sum.
Generated rolling-origin probabilistic forecasts across horizons of 1, 3, 6, and 12 months using covariance-weighted reconciliation (MinT-W) and nonnegative post-processing (NonNeg).
Assessed forecast performance using the Continuous Ranked Probability Score (CRPS), Logarithmic Score (LogS), and Diebold–Mariano tests with Holm adjustment in a deterministic pipeline.
Operationalized established reconciliation estimators into a transparent, deterministic pipeline that guarantees mathematical coherence and nonnegativity across hierarchical series.
Demonstrated statistically rigorous probabilistic evaluation across multi-month horizons using proper scoring rules, fully regenerable via a standardized reproduction package.

Cite This Study

Do Yeon Kim (2026) studied this question.

synapsesocial.com/papers/6a9bd4046b95aff0620eb469 https://doi.org/10.1177/18747655261484249

View Full Paper

Ask AI

Bookmark