Stratification by ancillary data in multisource forest inventories employing <i>k</i>-nearest-neighbour estimation
Bibliographic record
Abstract
The Finnish multisource national forest inventory (MS-NFI) utilizes optical area satellite images and digital maps in addition to field plot data to produce georeferenced information, thematic maps, and small-area statistics. In the early version, forestry land (FRYL) was taken directly from the numerical map data. Such data may be outdated and can contain significant errors, for example, the FRYL area is typically overestimated and the mean volume is underestimated. A statistical calibration method has been introduced to reduce the map errors on multisource forest resource estimates. It is based on large-area estimates of map errors, a confusion matrix among land-use classes of the field sample plots, and corresponding map information. The method has some drawbacks: calculations are more complicated than in the original MS-NFI and some field plots may have negative expansion factors. The paper presents a new stratified MS-NFI method to reduce the effect of inaccurate map data on the forest-resource estimates. In this method, the k-nearest-neighbour (k-NN) estimation is applied by strata. All the field plots within each map stratum, independently of their land-use classification by field crew, are used to estimate the areas of land-use classes and forest variables of that stratum. The method was tested on two large areas containing three Landsat 5 TM scenes and field-inventory data from the ninth NFI. The stratified MS-NFI is essentially a different estimation method compared with the calibrated MS-NFI, which calibrates the MS-NFI estimates more or less systematically in one direction. The stratified MS-NFI was found to be statistically simpler and there were fewer significant errors in the estimates than in the calibrated MS-NFI.
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 | 0.004 | 0.008 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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".