RandomForest classifier newClassifier Cross Validation Number of classes = 2 class 0.0: random_points_66_dry_20190612 accuracy = 0.9581 precision = 0.9484 correlation = 0.9196 errorRate = 0.0419 TruePositives = 1213.0000 FalsePositives = 66.0000 TrueNegatives = 1186.0000 FalseNegatives = 39.0000 class 1.0: random_points_66_wet_20190612 accuracy = 0.9581 precision = 0.9682 correlation = 0.9196 errorRate = 0.0419 TruePositives = 1186.0000 FalsePositives = 39.0000 TrueNegatives = 1213.0000 FalseNegatives = 66.0000 Using Testing dataset, % correct predictions = 95.8067 Total samples = 5010 RMSE = 0.2047752605864361 Bias = -0.010782747603833853 Distribution: class 0.0: random_points_66_dry_20190612 2505 (50.0000%) class 1.0: random_points_66_wet_20190612 2505 (50.0000%) Testing feature importance score: Each feature is perturbed 3 times and the % correct predictions are averaged The importance score is the original % correct prediction - average rank 1 feature 2 : VV score: tp=0.2129 accuracy=0.2129 precision=0.1552 correlation=0.3553 errorRate=-0.2129 cost=-0.2227 GainRatio = 0.2683 rank 2 feature 7 : VV+VH score: tp=0.1187 accuracy=0.1187 precision=0.0993 correlation=0.2093 errorRate=-0.1187 cost=-0.1239 GainRatio = 0.2761 rank 3 feature 1 : VH score: tp=0.0407 accuracy=0.0407 precision=0.0391 correlation=0.0764 errorRate=-0.0407 cost=-0.0426 GainRatio = 0.2515 rank 4 feature 8 : VV-VH score: tp=0.0369 accuracy=0.0369 precision=0.0365 correlation=0.0693 errorRate=-0.0369 cost=-0.0390 GainRatio = 0.0813 rank 5 feature 3 : VVrVH score: tp=0.0153 accuracy=0.0153 precision=0.0152 correlation=0.0294 errorRate=-0.0153 cost=-0.0159 GainRatio = 0.1582 rank 6 feature 4 : NVVI score: tp=0.0112 accuracy=0.0112 precision=0.0108 correlation=0.0216 errorRate=-0.0112 cost=-0.0114 GainRatio = 0.0850 rank 7 feature 5 : NDPI score: tp=0.0108 accuracy=0.0108 precision=0.0105 correlation=0.0208 errorRate=-0.0108 cost=-0.0110 GainRatio = 0.0850 rank 8 feature 6 : ndhi score: tp=0.0072 accuracy=0.0072 precision=0.0069 correlation=0.0140 errorRate=-0.0072 cost=-0.0073 GainRatio = 0.0850