Optimal combinations of data, classifiers, and sampling methods for accurate characterizations of deforestation
Bibliographic record
Abstract
AbstractThere are increasingly more choices from a complex of data resources, classification algorithms, and methods of training sample selections. To increase the repeatability of digital classifications of remotely sensed data with consistently high accuracy, it is essential to use optimal classification options or factors. In this paper, two temporal sets of Landsat thematic mapper (TM) data, three classifiers and three approaches of training sample selections were tested for mapping deforestation. The use of these different factors can have significant effects on classification accuracy. The mixed effects of the three factors can also magnify the variations of classification accuracy. The use of bi-temporal data, a spatial‐spectral classifier, and hybrid training samples results in steadily higher classification accuracy than the combination of uni-temporal data, a spectral classifier, and image training samples. For the purpose of characterizing managed forest lands, even a small increase in overall accuracy of image classification is important because it may represent a large decrease in the variations of the producer's and user's accuracy, which in turn can reduce the uncertainties of area measurements for forest coverage.Il y a de plus en plus de possibilités au niveau du choix des sources de données, des algorithmes de classification et des méthodes de sélection de sites d'entraînement. Pour accroître la répétitivité des classifications numériques des données de télédétection avec une précision constante élevée, il est essentiel d'utiliser les options ou les facteurs optimaux de classification. Dans cet article, deux ensembles de données temporelles Landsat TM, trois classificateurs et trois approches de sélection de sites d'entraînement ont été testés pour la cartographie de la déforestation. L'utilisation de ces différents facteurs peut avoir des effets significatifs sur la précision de classification. Les effets combinés de ces trois facteurs peuvent aussi accroître les variations de la précision de classification. L'utilisation de données bi-temporelles, d'un classificateur spatial‐spectral et de sites d'entraînement hybrides accroît de façon constante la précision de classification comparativement à la combinaison de données uni-temporelles, d'un classificateur spectral et de sites d'entraînement. Dans le contexte de la caractérisation des terres forestières sous gestion, même un faible taux d'accroissement dans la précision globale de classification de l'image est important parce qu'il peut correspondre à une forte diminution dans les variations de précision des producteurs et de l'usager, qui en retour, peut réduire les incertitudes dans les mesures de surface de la couverture forestière.[Traduit par la Rédaction]
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 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 teacher head, 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".