Applying 2022-01-21 shrubland models to all lidar coverages.
This report is a companion to Shrubland 1.0.2: Balanced Diet. It describes the same models as that document.
The logistic ensemble model was used to predict shrubland in all areas with LiDAR-derived CHMs. Predictions were restricted to only pixels of vegetated LCPRI classes (excluding developed, barren, ice/snow, and water). The predicted probabilities were then classified using four thresholds, documented in the model report document, chosen to target certain values of specificity in order to improve positive predictive value.
These results pool training, validation, and test pixels with the remainder of the state. As only 00.02% of pixels were used in any of these roles, these results are for all practical purposes equal to out-of-bag accuracy.
Threshold: 0.489
Confusion Matrix and Statistics
1 0
1 1807278 17342880
0 300211 62427820
Accuracy : 0.7845
95% CI : (0.7844, 0.7846)
No Information Rate : 0.9743
P-Value [Acc > NIR] : 1
Kappa : 0.1297
Mcnemar's Test P-Value : <2e-16
Sensitivity : 0.85755
Specificity : 0.78259
Pos Pred Value : 0.09437
Neg Pred Value : 0.99521
Prevalence : 0.02574
Detection Rate : 0.02207
Detection Prevalence : 0.23389
Balanced Accuracy : 0.82007
'Positive' Class : 1
Threshold: 0.755
Confusion Matrix and Statistics
1 0
1 1451903 8282306
0 655586 71488394
Accuracy : 0.8908
95% CI : (0.8908, 0.8909)
No Information Rate : 0.9743
P-Value [Acc > NIR] : 1
Kappa : 0.2119
Mcnemar's Test P-Value : <2e-16
Sensitivity : 0.68893
Specificity : 0.89617
Pos Pred Value : 0.14915
Neg Pred Value : 0.99091
Prevalence : 0.02574
Detection Rate : 0.01773
Detection Prevalence : 0.11889
Balanced Accuracy : 0.79255
'Positive' Class : 1
Threshold: 0.840
Confusion Matrix and Statistics
1 0
1 1083602 3875692
0 1023887 75895008
Accuracy : 0.9402
95% CI : (0.9401, 0.9402)
No Information Rate : 0.9743
P-Value [Acc > NIR] : 1
Kappa : 0.2807
Mcnemar's Test P-Value : <2e-16
Sensitivity : 0.51417
Specificity : 0.95141
Pos Pred Value : 0.21850
Neg Pred Value : 0.98669
Prevalence : 0.02574
Detection Rate : 0.01323
Detection Prevalence : 0.06057
Balanced Accuracy : 0.73279
'Positive' Class : 1
Threshold: 0.907
Confusion Matrix and Statistics
1 0
1 519993 861375
0 1587496 78909325
Accuracy : 0.9701
95% CI : (0.9701, 0.9701)
No Information Rate : 0.9743
P-Value [Acc > NIR] : 1
Kappa : 0.2835
Mcnemar's Test P-Value : <2e-16
Sensitivity : 0.246736
Specificity : 0.989202
Pos Pred Value : 0.376433
Neg Pred Value : 0.980279
Prevalence : 0.025739
Detection Rate : 0.006351
Detection Prevalence : 0.016871
Balanced Accuracy : 0.617969
'Positive' Class : 1
Raster files can be downloaded from GitHub at this link.
Files ending in classified have been classified into four categories:
Files not ending in classified represent the binary shrubland maps (1 = shrub) using the threshold specified above.
The probability map is too large to share easily; I’ll be adding it to Labrador in the near future.
If you see mistakes or want to suggest changes, please create an issue on the source repository.
For attribution, please cite this work as
Mahoney (2022, Jan. 22). CAFRI Labs: Shrubland 1.0.2: Statewide Prediction Accuracy. Retrieved from https://cafri-labs.github.io/acceptable-growing-stock/posts/shrubland-102-statewide-prediction-accuracy/
BibTeX citation
@misc{mahoney2022shrubland,
author = {Mahoney, Mike},
title = {CAFRI Labs: Shrubland 1.0.2: Statewide Prediction Accuracy},
url = {https://cafri-labs.github.io/acceptable-growing-stock/posts/shrubland-102-statewide-prediction-accuracy/},
year = {2022}
}