3. Random Forest:
with 10 trees:
on all data:
Correctly Classified Instances 24 70.5882 %
Incorrectly Classified Instances 10 29.4118 %
Kappa statistic 0.1981
Mean absolute error 0.4095
Root mean squared error 0.4614
Relative absolute error 97.2302 %
Root relative squared error 101.111 %
Total Number of Instances 34
=== Detailed Accuracy By Class ===
TP Rate FP Rate Precision Recall F-Measure ROC Area Class
0.3 0.125 0.5 0.3 0.375 0.613 AR
0.875 0.7 0.75 0.875 0.808 0.613 NR
Weighted Avg. 0.706 0.531 0.676 0.706 0.68 0.613
=== Confusion Matrix ===
a b <-- classified as
3 7 | a = AR
3 21 | b = NR
on the subset of peptides with p value < 0.1:
Correctly Classified Instances 28 82.3529 %
Incorrectly Classified Instances 6 17.6471 %
Kappa statistic 0.5984
Mean absolute error 0.3
Root mean squared error 0.3835
Relative absolute error 71.2381 %
Root relative squared error 84.0423 %
Total Number of Instances 34
=== Detailed Accuracy By Class ===
TP Rate FP Rate Precision Recall F-Measure ROC Area Class
0.8 0.167 0.667 0.8 0.727 0.871 AR
0.833 0.2 0.909 0.833 0.87 0.871 NR
Weighted Avg. 0.824 0.19 0.838 0.824 0.828 0.871
=== Confusion Matrix ===
a b <-- classified as
8 2 | a = AR
4 20 | b = NR
on the choice of 4 peptides:
Correctly Classified Instances 25 73.5294 %
Incorrectly Classified Instances 9 26.4706 %
Kappa statistic 0.4138
Mean absolute error 0.3176
Root mean squared error 0.4144
Relative absolute error 75.4286 %
Root relative squared error 90.828 %
Total Number of Instances 34
=== Detailed Accuracy By Class ===
TP Rate FP Rate Precision Recall F-Measure ROC Area Class
0.7 0.25 0.538 0.7 0.609 0.8 AR
0.75 0.3 0.857 0.75 0.8 0.8 NR
Weighted Avg. 0.735 0.285 0.763 0.735 0.744 0.8
=== Confusion Matrix ===
a b <-- classified as
7 3 | a = AR
6 18 | b = NR
with 5 trees:
on all data:
Correctly Classified Instances 23 67.6471 %
Incorrectly Classified Instances 11 32.3529 %
Kappa statistic 0.0878
Mean absolute error 0.3804
Root mean squared error 0.4571
Relative absolute error 90.3293 %
Root relative squared error 100.1854 %
Total Number of Instances 34
=== Detailed Accuracy By Class ===
TP Rate FP Rate Precision Recall F-Measure ROC Area Class
0.2 0.125 0.4 0.2 0.267 0.677 AR
0.875 0.8 0.724 0.875 0.792 0.677 NR
Weighted Avg. 0.676 0.601 0.629 0.676 0.638 0.677
=== Confusion Matrix ===
a b <-- classified as
2 8 | a = AR
3 21 | b = NR
on the subset of data for which the p value is <0.1:
Correctly Classified Instances 29 85.2941 %
Incorrectly Classified Instances 5 14.7059 %
Kappa statistic 0.6559
Mean absolute error 0.2353
Root mean squared error 0.3662
Relative absolute error 55.873 %
Root relative squared error 80.2593 %
Total Number of Instances 34
=== Detailed Accuracy By Class ===
TP Rate FP Rate Precision Recall F-Measure ROC Area Class
0.8 0.125 0.727 0.8 0.762 0.858 AR
0.875 0.2 0.913 0.875 0.894 0.858 NR
Weighted Avg. 0.853 0.178 0.858 0.853 0.855 0.858
=== Confusion Matrix ===
a b <-- classified as
8 2 | a = AR
3 21 | b = NR
on the choice of 4 peptides:
Correctly Classified Instances 24 70.5882 %
Incorrectly Classified Instances 10 29.4118 %
Kappa statistic 0.3657
Mean absolute error 0.3059
Root mean squared error 0.4366
Relative absolute error 72.6349 %
Root relative squared error 95.6755 %
Total Number of Instances 34
=== Detailed Accuracy By Class ===
TP Rate FP Rate Precision Recall F-Measure ROC Area Class
0.7 0.292 0.5 0.7 0.583 0.819 AR
0.708 0.3 0.85 0.708 0.773 0.819 NR
Weighted Avg. 0.706 0.298 0.747 0.706 0.717 0.819
=== Confusion Matrix ===
a b <-- classified as
7 3 | a = AR
7 17 | b = NR
* playing with the number of features or the maximum depth of the tree did not help with increasing the accuracy.
No comments:
Post a Comment