Correlation within SNCF Administrative Regions among Track Segment Maintenance Cost Equation Residuals of a Countrywide Model
Bibliographic record
Abstract
Using as reference a recent France-wide model of rail infrastructure maintenance cost where regression model residuals associated to track segments are assumed to be similarly correlated among themselves within the 23 administrative regions of the national firm (SNCF), we attempt to explain the presence of the strong positive and stationary correlation coefficient estimated in this manner and to probe the extent to which the assumption of a common correlation coefficient across administrative regions might be refined and interpreted.We first find that country-wide within-region correlation among residuals is not weakened if regional dummy variables are added to the model in the hope of finding interpretative clues, notably of the presence of climate effects unaccounted for in the model specification.The implicit regionspecific weather effects indirectly so represented turn out to be extremely weak, if present at all, and do not affect the strength of extant within-region correlation or the need to make sense of it.We then explore differences in correlation coefficient estimates among regions and show that, within the reference maintenance cost model, two large geographic groupings of regions, each comprising in the East or in the North about 15% of total available track segments, in fact have own residuals that are uncorrelated among themselves, in contrast to the more numerous 70% of segment residuals remaining in the rest of regions, which as a group remain robustly and positively correlated, and in a stationary manner.As the two groupings with uncorrelated residuals closely match the networks of regional firms merged into the SNCF conglomerate in 1938, we hypothesize, faute de mieux and in the absence for the moment of refined local climate variables to pursue unpromising missing variable tests, that within-firm accounting traditions might have survived centralized management control of the assignment of track surveillance, maintenance and repair invoices to track sections centrally defined for accounting management purposes.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.005 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.007 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".