|
Inner Ear: SVM Linear Kernel
|
Middle Ear: Logistic Regression
|
| |
Predicted Label
| |
Predicted Label
|
|
Actual Label
|
Normal
|
Abnormal
|
Actual Label
|
Normal
|
Abnormal
|
|
Normal
|
101
|
7
|
Normal
|
87
|
0
|
|
Abnormal
|
7
|
31
|
Abnormal
|
15
|
44
|
|
Outer Ear: SVM Linear Kernel
|
Mastoid: Decision Tree
|
| |
Predicted Label
| |
Predicted Label
|
|
Actual Label
|
Normal
|
Abnormal
|
Actual Label
|
Normal
|
Abnormal
|
|
Normal
|
124
|
1
|
Normal
|
83
|
9
|
|
Abnormal
|
9
|
12
|
Abnormal
|
17
|
37
|
- Test set confusion matrices for best learning algorithm for each ear region. The confusion matrices provide the true and false counts for normal and abnormal documents as labeled by the classification algorithm
