Evaluation Results
Last version of FIA models: 2022-04-11 Final Countdown
Change Summary
- This version adds 65 indices and their respective 1 year deltas (130 predictors) computed through awesome spectral indices library.
| RF (ranger) | GBM (LightGBM) | SVM (kernlab) | Ensemble (model weighted) | Ensemble (RMSE weighted) | |
|---|---|---|---|---|---|
| %RMSE | 39.729 | 40.433 | 39.233 | 38.758 | 39.174 |
| RMSE | 54.577 | 55.545 | 53.897 | 53.244 | 53.816 |
| MAE | 42.941 | 43.655 | 42.093 | 41.869 | 42.259 |
| MBE | 1.578 | 1.225 | -1.228 | 1.179 | 0.506 |
| R2 | 0.362 | 0.339 | 0.374 | 0.389 | 0.382 |
AGB Distribution
summary(bind_rows(training, testing)$agb_mgha)
Min. 1st Qu. Median Mean 3rd Qu. Max.
0.00 89.51 134.78 136.82 179.32 425.00 
Metadata
Ensembles
- RMSE-weighted model weights:
- Linear model weights:
Call:
lm(formula = agb_mgha ~ rf_pred * lgb_pred * svm_pred, data = k_fold_preds)
Residuals:
Min 1Q Median 3Q Max
-158.666 -34.994 -3.824 30.331 220.747
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) -6.073e-01 2.551e+01 -0.024 0.9810
rf_pred -5.753e-01 4.238e-01 -1.358 0.1748
lgb_pred 3.150e-01 5.250e-01 0.600 0.5486
svm_pred 7.263e-01 3.561e-01 2.040 0.0416 *
rf_pred:lgb_pred 4.386e-03 3.060e-03 1.433 0.1520
rf_pred:svm_pred 4.379e-03 2.903e-03 1.509 0.1316
lgb_pred:svm_pred -1.537e-03 3.671e-03 -0.419 0.6754
rf_pred:lgb_pred:svm_pred -2.279e-05 1.460e-05 -1.561 0.1188
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Residual standard error: 52.65 on 1407 degrees of freedom
Multiple R-squared: 0.4093, Adjusted R-squared: 0.4063
F-statistic: 139.2 on 7 and 1407 DF, p-value: < 2.2e-16Coverages
- 0 coverages:
\(n\) and \(p\)
- 2028 observations
- 1419 training
- 609 testing
- 156 predictors
- aweinsh, delta_aweinsh, aweish, delta_aweish, bcc, delta_bcc, dvi, delta_dvi, embi, delta_embi, exgr, delta_exgr, exr, delta_exr, fcvi, delta_fcvi, gari, delta_gari, gbndvi, delta_gbndvi, gli, delta_gli, gndvi, delta_gndvi, gosavi, delta_gosavi, grndvi, delta_grndvi, gvmi, delta_gvmi, ikaw, delta_ikaw, ipvi, delta_ipvi, lswi, delta_lswi, mag, mbi, delta_mbi, mcari2, delta_mcari2, mlswi26, delta_mlswi26, mlswi27, delta_mlswi27, mndwi, delta_mndwi, mrbvi, delta_mrbvi, msavi, delta_msavi, msi, delta_msi, msr, delta_msr, mtvi2, delta_mtvi2, muwir, delta_muwir, nbr, delta_nbr, nbr2, delta_nbr2, nbsims, delta_nbsims, nddi, delta_nddi, ndsi, delta_ndsi, ndvi, delta_ndvi, ndvimndwi, delta_ndvimndwi, ndwi, delta_ndwi, ngrdi, delta_ngrdi, nirv, delta_nirv, nli, delta_nli, nmdi, delta_nmdi, normg, delta_normg, normnir, delta_normnir, normr, delta_normr, nrfig, delta_nrfig, nsds, delta_nsds, nwi, delta_nwi, pisi, delta_pisi, rcc, delta_rcc, rdvi, delta_rdvi, rgbvi, delta_rgbvi, rgri, delta_rgri, s3, delta_s3, si, delta_si, sr, delta_sr, sr2, delta_sr2, swi, delta_swi, tcb, delta_tcb, tcg, delta_tcg, tcw, delta_tcw, tdvi, delta_tdvi, trivi, delta_trivi, tvi, delta_tvi, ui, delta_ui, vgnirbi, delta_vgnirbi, vig, delta_vig, vrnirbi, delta_vrnirbi, wi1, delta_wi1, wi2, delta_wi2, yod, precip, tmin, tmax, dem, twi, slope, aspect, chm, lcpri4, lcpri2, lcpri6, lcpri3, lcsec3, lcsec1, lcsec2, lcsec4
Component Models
- Tuning used 5-fold CV
- Final hyperparameters:
Random forest:
$num.trees
[1] 3500
$mtry
[1] 36
$min.node.size
[1] 4
$sample.fraction
[1] 1
$replace
[1] FALSE
$formula
agb_mgha ~ .LGB:
$nrounds
[1] 50
$params
$params$learning_rate
[1] 0.05
$params$num_leaves
[1] 36
$params$max_depth
[1] -1
$params$extra_trees
[1] TRUE
$params$min_data_in_leaf
[1] 3
$params$bagging_fraction
[1] 0.5
$params$bagging_freq
[1] 10
$params$feature_fraction
[1] 0.8
$params$min_data_in_bin
[1] 13
$params$lambda_l1
[1] 8
$params$lambda_l2
[1] 0
$params$force_col_wise
[1] TRUESVM:
$x
agb_mgha ~ .
$kernel
[1] "laplacedot"
$type
[1] "eps-svr"
$kpar
$kpar$sigma
[1] 0.015625
$C
[1] 8
$epsilon
[1] 0.125
$nu
[1] 0.2