RandomForest classifier newClassifier Cross Validation Number of classes = 7 class 0.0: BR26 accuracy = 0.9968 precision = 0.9970 correlation = 0.9897 errorRate = 0.0032 TruePositives = 4038.0000 FalsePositives = 12.0000 TrueNegatives = 946.0000 FalseNegatives = 4.0000 class 5.0: null accuracy = 0.9952 precision = 0.9742 correlation = 0.9595 errorRate = 0.0048 TruePositives = 302.0000 FalsePositives = 8.0000 TrueNegatives = 4674.0000 FalseNegatives = 16.0000 class 10.0: null accuracy = 0.9976 precision = 0.9646 correlation = 0.9721 errorRate = 0.0024 TruePositives = 218.0000 FalsePositives = 8.0000 TrueNegatives = 4770.0000 FalseNegatives = 4.0000 class 15.0: null accuracy = 0.9972 precision = 0.9858 correlation = 0.9788 errorRate = 0.0028 TruePositives = 348.0000 FalsePositives = 5.0000 TrueNegatives = 4638.0000 FalseNegatives = 9.0000 class 20.0: null accuracy = 0.9976 precision = 0.8909 correlation = 0.8900 errorRate = 0.0024 TruePositives = 49.0000 FalsePositives = 6.0000 TrueNegatives = 4939.0000 FalseNegatives = 6.0000 class 25.0: null accuracy = 0.9986 precision = 0.0000 correlation = 0.0007 errorRate = 0.0014 TruePositives = 0.0000 FalsePositives = 3.0000 TrueNegatives = 4993.0000 FalseNegatives = 4.0000 class 30.0: null accuracy = 0.9990 precision = 0.0000 correlation = 0.0005 errorRate = 0.0010 TruePositives = 0.0000 FalsePositives = 3.0000 TrueNegatives = 4995.0000 FalseNegatives = 2.0000 Using Testing dataset, % correct predictions = 99.1000 Total samples = 10000 RMSE = 0.4743416490252569 Bias = -0.01100000000000012 Distribution: class 0.0: BR26 8106 (81.0600%) class 5.0: null 698 (6.9800%) class 10.0: null 317 (3.1700%) class 15.0: null 485 (4.8500%) class 20.0: null 265 (2.6500%) class 25.0: null 122 (1.2200%) class 30.0: null 7 (0.0700%) 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 16 : NDSI37 score: tp=0.0103 accuracy=0.0030 precision=0.0087 correlation=0.0215 errorRate=-0.0030 cost=-0.0092 GainRatio = 0.2883 rank 2 feature 12 : NDSI36 score: tp=0.0093 accuracy=0.0026 precision=0.0087 correlation=0.0153 errorRate=-0.0026 cost=-0.0092 GainRatio = 0.2994 rank 3 feature 6 : B7 score: tp=0.0063 accuracy=0.0018 precision=0.0045 correlation=0.0130 errorRate=-0.0018 cost=-0.0046 GainRatio = 0.1218 rank 4 feature 1 : B2 score: tp=0.0049 accuracy=0.0014 precision=0.0083 correlation=0.0154 errorRate=-0.0014 cost=-0.0086 GainRatio = 0.5267 rank 5 feature 5 : B6 score: tp=0.0042 accuracy=0.0012 precision=0.0024 correlation=0.0093 errorRate=-0.0012 cost=-0.0024 GainRatio = 0.1357 rank 6 feature 4 : B5 score: tp=0.0035 accuracy=0.0010 precision=0.0071 correlation=0.0141 errorRate=-0.0010 cost=-0.0073 GainRatio = 0.2989 rank 7 feature 3 : B4 score: tp=0.0027 accuracy=0.0008 precision=0.0081 correlation=0.0537 errorRate=-0.0008 cost=-0.0086 GainRatio = 0.3765 rank 8 feature 8 : BR36 score: tp=0.0023 accuracy=0.0007 precision=0.0069 correlation=0.0059 errorRate=-0.0007 cost=-0.0072 GainRatio = 0.7863 rank 9 feature 7 : BR27 score: tp=0.0017 accuracy=0.0005 precision=0.0060 correlation=0.0138 errorRate=-0.0005 cost=-0.0063 GainRatio = 0.7957 rank 10 feature 14 : 56 score: tp=0.0015 accuracy=0.0004 precision=0.0073 correlation=0.0079 errorRate=-0.0004 cost=-0.0078 GainRatio = 0.7790 rank 11 feature 15 : 57 score: tp=0.0009 accuracy=0.0002 precision=0.0045 correlation=0.0036 errorRate=-0.0002 cost=-0.0048 GainRatio = 0.7631 rank 12 feature 9 : BR37 score: tp=0.0008 accuracy=0.0002 precision=0.0041 correlation=0.0052 errorRate=-0.0002 cost=-0.0044 GainRatio = 0.7601 rank 13 feature 13 : elevation score: tp=0.0007 accuracy=0.0002 precision=-0.0021 correlation=0.0002 errorRate=-0.0002 cost=0.0022 GainRatio = 0.1077 rank 14 feature 10 : BR46 score: tp=0.0006 accuracy=0.0002 precision=0.0015 correlation=0.0030 errorRate=-0.0002 cost=-0.0016 GainRatio = 0.7868 Warning: rank <= featureBandList.length