Testing spatial out-of-sample area of influence for grain forecasting models
Bibliographic record
Abstract
Abstract We examine the factors that determine if a grain forecasting model fit to one region can be transferred to another region. Prior research has proposed examining the area of applicability (AoA) of a model based on structurally similar characteristics in the Earth Observation predictors and weights based on the model derived feature importance. We expand on and evaluate this approach in the context of grain yield forecasting in Sub-Saharan Africa. Specifically, we evaluate an AoA methodology established for generating raster surfaces and apply it to vector supported grain data. We fit a series of ensemble tree models both within single countries and across multiple sets of countries and then test those models in countries excluded from the training set. We then calculate and decompose AoA measures and examine several different performance metrics. We find that the spatial transfer accuracy does not vary across season but does vary by average rainfall and across high, medium, and low yielding regions. In general, areas with higher yields and medium to high average rainfall tend to have higher accuracy for both model training and transfer. Finally, we find that fitting models with multiple countries provides more accurate out-of-sample estimates when compared to models fitted to a single country.
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 | 0.035 | 0.104 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.002 | 0.003 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
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".