Results for classifiers learnt from 20 initial configurations.

========
TABLE II
========

model type           accuracy    cross-validation     sensitivity    specificity
SVM-linear            52.09%         50.04%              62.18%         42.03%
SVM-polynomial        58.08%         51.01%              61.68%         54.48%
SVM-RBF               95.73%         94.77%              99.78%         91.68%



=========
TABLE III
=========

number of mutants        effective mutants         accuracy           cross-validation      
300                             32                   85.75%               81.01%
400                             52                   90.11%               89.98%
500                             71                   94.01%               95.04%
600                             91                   95.73%               95.09%
700                             109                  95.73%               94.77%


========
TABLE IV
========
#time interval       accuracy     cross-validation
100                   94.54%           93.25%
150                   94.14%           92.14%
200                   95.79%           94.45%
250                   95.73%           94.77%
300                   94.98%           94.31%
=======
TABLE V
=======

attack #             detected         accuracy
1                      yes              90.16%
2                      yes              94.09%
3                      yes              71.11%               
4                      yes              98.14%         
5                      yes              95.12%
6                      yes              91.09%  
7                      yes              76.98%      
8                      yes              94.14%      
9                      yes              95.52%
10                     yes              99.11%
11                     yes              90.19%
12                     yes              95.16%      
13                     yes              98.97%
14                     yes              95.47%
15                     yes              94.79%
========
TABLE VI
========
attack stage   # effective mutants        # detected     accuracy (detected)    accuracy (all)
PLC1                    8                      5               99.18%               74.84%    
PLC3                   20                     17               98.19%               91.92%
PLC4                    5                      4               99.09%               90.09%
PLC5                    2                      2               98.13%               98.13%                     
PLC6                    5                      5               98.97%               99.97%
summary                40                     33               98.41%               90.09%