gkj115_SupplementaryTableS2 by shitingting

VIEWS: 11 PAGES: 8

									Table S2. Classifcation results (shown as confusion matrices) for the Wnt-signaling dataset
(number of positives = 16, number of negatives = 500)
The classification is performed with a 10 fold cross-validation using Weka's classification scheme's




                                                                                  predicted negative
                                                            predicted positive
                                known positive          TP                       FN
                                known negative          FP                       TN



                                                             Transfac 8.4                                            phyloFACTS              JASPAR CORE
Classification algorithm Atrribute selection algorithm Confusion Matrix F-mea-                                 Confusion Matrix F-mea-   Confusion Matrix
                                                                        sure                                                    sure
Logistic

Logistic                 all attributes                 7                                 9            0.4     4          12   0.2       11         5
                                                        12                              488                    20        480             17       483

Logistic                 InfoGain/Ranker T = 0.005      10                                6            0.556   3          13   0.273     8          8
                                                        10                              490                    3         497             3        497

Logistic                 InfoGain/Ranker T = 0.010      10                                6            0.556   3          13   0.273     8          8
                                                        10                              490                    3         497             3        497

Logistic                 InfoGain/Ranker T = 0.015      9                                 7            0.474   2          14   0.2       9          7
                                                        13                              487                    2         498             4        496

Logistic                 InfoGain/Ranker T = 0.020      9                                 7            0.529   3          13   0.286     9          7
                                                        9                               491                    2         498             2        498

Logistic                 InfoGain/Ranker T = 0.025      8                                 8            0.615   3          13   0.286     9          7
                                                        2                               498                    2         498             1        499

Logistic                 InfoGain/Ranker T = 0.030      9                                 7            0.692   3          13   0.286     8          8
                                                        1                               499                    2         498             0        500

Logistic                 CfsSubsetEval/BF               10                                6            0.741   3          13   0.273     9          7
                                                        1                               499                    3         497             2        498

SMO (Support Vector Machine)

SMO C=1                  all attributes                 5                                11            0.435   2          14   0.174     5         11
                                                        2                               498                    5         495             0        500

SMO C=10                 all attributes                 5                                11            0.435   2          14   0.148     8          8
                                                        2                               498                    9         491             3        497

SMO C=100                all attributes                 5                                11            0.435   2          14   0.148     8          8
                                                        2                               498                    9         491             3        497

SMO C=10.000             all attributes                 5                                11            0.435   2          14   0.148     8          8
                                                        2                               498                    9         491             3        497

SMO C=1.000.000          all attributes                 5                                11            0.435   2          14   0.148     8          8
                                                        2                               498                    9         491             3        497

SMO C=1                  InfoGain/Ranker T = 0.005      4                                12            0.381   0          16   0         4         12
                                                        1                               499                    0         500             0        500

SMO C=10                 InfoGain/Ranker T = 0.005      9                                 7            0.667   1          15   0.118     8          8
                                                        2                               498                    0         500             2        498
SMO C=100         InfoGain/Ranker T = 0.005   10     6   0.588   1    15   0.111   8     8
                                              8    492           1   499           2   498

SMO C=10.000      InfoGain/Ranker T = 0.005   9      7   0.462   1    15   0.111   8     8
                                              14   486           1   499           3   497

SMO C=1.000.000   InfoGain/Ranker T = 0.005   10     6   0.5     1    15   0.111   8     8
                                              14   486           1   499           3   497

SMO C=1           InfoGain/Ranker T = 0.010   4     12   0.381   0    16   0       4    12
                                              1    499           0   500           0   500

SMO C=10          InfoGain/Ranker T = 0.010   9      7   0.667   2    14   0.211   8     8
                                              2    498           1   499           2   498

SMO C=100         InfoGain/Ranker T = 0.010   10     6   0.588   2    14   0.211   8     8
                                              8    492           1   499           2   498

SMO C=10.000      InfoGain/Ranker T = 0.010   9      7   0.462   2    14   0.211   8     8
                                              14   486           1   499           3   497

SMO C=1.000.000   InfoGain/Ranker T = 0.010   10     6   0.5     2    14   0.211   8     8
                                              14   486           1   499           3   497

SMO C=1           InfoGain/Ranker T = 0.015   5     11   0.476   0    16   0       4    12
                                              0    500           0   500           0   500

SMO C=10          InfoGain/Ranker T = 0.015   8      8   0.667   0    16   0       8     8
                                              0    500           1   499           1   499

SMO C=100         InfoGain/Ranker T = 0.015   8      8   0.516   0    16   0       8     8
                                              7    493           1   499           1   499

SMO C=10.000      InfoGain/Ranker T = 0.015   9      7   0.5     0    16   0       9     7
                                              11   489           1   499           3   497

SMO C=1.000.000   InfoGain/Ranker T = 0.015   9      7   0.5     0    16   0       9     7
                                              11   489           1   499           3   497

SMO C=1           InfoGain/Ranker T = 0.020   4     12   0.4     0    16   0       5    11
                                              0    500           0   500           0   500

SMO C=10          InfoGain/Ranker T = 0.020   8      8   0.667   0    16   0       8     8
                                              0    500           0   500           0   500

SMO C=100         InfoGain/Ranker T = 0.020   8      8   0.667   0    16   0       8     8
                                              0    500           0   500           0   500

SMO C=10.000      InfoGain/Ranker T = 0.020   8      8   0.571   0    16   0       9     7
                                              4    496           0   500           0   500

SMO C=1.000.000   InfoGain/Ranker T = 0.020   8      8   0.552   0    16   0       9     7
                                              5    495           0   500           0   500

SMO C=1           InfoGain/Ranker T = 0.025   4     12   0.4     0    16   0       5    11
                                              0    500           0   500           0   500

SMO C=10          InfoGain/Ranker T = 0.025   8      8   0.667   0    16   0       8     8
                                              0    500           0   500           0   500

SMO C=100         InfoGain/Ranker T = 0.025   8      8   0.667   0    16   0       9     7
                                              0    500           0   500           0   500

SMO C=10.000      InfoGain/Ranker T = 0.025   8      8   0.64    0    16   0       9     7
                                              1    499           0   500           0   500

SMO C=1.000.000   InfoGain/Ranker T = 0.025   8      8   0.64    0    16   0       9     7
                                              1    499           0   500           0   500

SMO C=1           CfsSubsetEval/BF            5     11   0.476   0    16   0       5    11
                                              0    500           0   500           0   500

SMO C=10          CfsSubsetEval/BF            8      8   0.667   0    16   0       8     8
                                              0    500           1   499           0   500
SMO C=100                 CfsSubsetEval/BF            9      7   0.72    0     16   0       8      8
                                                      0    500           1    499           1    499

SMO C=10.000              CfsSubsetEval/BF            8      8   0.64    0     16   0       8      8
                                                      1    499           1    499           1    499

SMO C=1.000.000           CfsSubsetEval/BF            8      8   0.64    0     16   0       8      8
                                                      1    499           1    499           1    499

Ibk (Nearest Neighbour)

IBk K=1                   all attributes              1     15   0.095   1     15   0.067   4     12
                                                      4    496           13   487           4    496

IBk K=3                   all attributes              1     15   0.118   0     16   0       1     15
                                                      0    500           0    500           1    499

IBk K=5                   all attributes              0     16   0       0     16   0       0     16
                                                      0    500           0    500           0    500

IBk K=1                   InfoGain/Ranker T = 0.005   0     16   0       4     12   0.32    5     11
                                                      6    494           5    495           7    493

IBk K=1                   InfoGain/Ranker T = 0.010   0     16   0       6     10   0.444   5     11
                                                      6    494           5    495           7    493

IBk K=1                   InfoGain/Ranker T = 0.015   1     15   0.077   6     10   0.48    7      9
                                                      9    491           3    497           4    496

IBk K=1                   InfoGain/Ranker T = 0.020   4     12   0.308   6     10   0.48    6     10
                                                      6    494           3    497           6    494

IBk K=1                   InfoGain/Ranker T = 0.025   5     11   0.323   6     10   0.48    9      7
                                                      10   490           3    497           3    497

IBk K=1                   InfoGain/Ranker T = 0.030   4     12   0.286   6     10   0.48    10     6
                                                      8    492           3    497           1    499

IBk K=1                   CfsSubsetEval/BF            7      9   0.5     7      9   0.519   6     10
                                                      5    495           4    496           6    494

NB

NB                        all attributes              9      7   0.184   6     10   0.19    12     4
                                                      73   427           41   459           88   412

NB                        InfoGain/Ranker T = 0.005   11     5   0.306   5     11   0.263   9      7
                                                      45   455           17   483           15   485

NB                        InfoGain/Ranker T = 0.010   11     5   0.306   5     11   0.263   9      7
                                                      45   455           17   483           15   485

NB                        InfoGain/Ranker T = 0.015   10     6   0.282   6     10   0.48    9      7
                                                      45   455           3    497           8    492

NB                        InfoGain/Ranker T = 0.020   11     5   0.314   7      9   0.538   9      7
                                                      43   457           3    497           7    493

NB                        InfoGain/Ranker T = 0.025   9      7   0.321   7      9   0.538   10     6
                                                      31   469           3    497           9    491

NB                        InfoGain/Ranker T = 0.030   9      7   0.367   7      9   0.538   9      7
                                                      24   476           3    497           5    495

NB                        CfsSubsetEval/BF            9      7   0.439   5     11   0.345   9      7
                                                      16   484           8    492           17   483

NBTree

NBTree                    all attributes              5     11   0.385   2     14   0.2     5     11
                                                      5    495           2    498           4    496

NBTree                    InfoGain/Ranker T = 0.005   6     10   0.462   0     16   0       8      8
                                                   4   496           4   496           1   499

NBTree                 InfoGain/Ranker T = 0.010   6    10   0.462   0    16   0       8     8
                                                   4   496           4   496           1   499

NBTree                 InfoGain/Ranker T = 0.015   5    11   0.4     0    16   0       8     8
                                                   4   496           5   495           0   500

NBTree                 InfoGain/Ranker T = 0.020   8     8   0.615   0    16   0       8     8
                                                   2   498           3   497           0   500

NBTree                 InfoGain/Ranker T = 0.025   8     8   0.667   0    16   0       8     8
                                                   0   500           3   497           0   500

NBTree                 InfoGain/Ranker T = 0.030   8     8   0.64    0    16   0       8     8
                                                   1   499           3   497           0   500

NBTree                 CfsSubsetEval/BF            7     9   0.583   0    16   0       8     8
                                                   1   499           4   496           0   500

MultiLayerPerceptron

MultiLayerPerceptron   all attributes              4    12   0.381   3    13   0.25    8     8
                                                   1   499           5   495           1   499

MultiLayerPerceptron   InfoGain/Ranker T = 0.005   8     8   0.593   2    14   0.2     9     7
                                                   3   497           2   498           5   495

MultiLayerPerceptron   InfoGain/Ranker T = 0.010   8     8   0.593   3    13   0.273   9     7
                                                   3   497           3   497           5   495

MultiLayerPerceptron   InfoGain/Ranker T = 0.015   8     8   0.615   3    13   0.3     8     8
                                                   2   498           1   499           4   496

MultiLayerPerceptron   InfoGain/Ranker T = 0.020   8     8   0.593   3    13   0.286   8     8
                                                   3   497           2   498           2   498

MultiLayerPerceptron   InfoGain/Ranker T = 0.025   7     9   0.538   3    13   0.286   8     8
                                                   3   497           2   498           0   500

MultiLayerPerceptron   InfoGain/Ranker T = 0.030   8     8   0.64    3    13   0.286   8     8
                                                   1   499           2   498           0   500

MultiLayerPerceptron   CfsSubsetEval/BF            8     8   0.571   3    13   0.286   8     8
                                                   4   496           2   498           2   498

J48

J48                    all attributes              9     7   0.621   3    13   0.3     6    10
                                                   4   496           1   499           3   497

J48                    InfoGain/Ranker T = 0.005   9     7   0.643   2    14   0.222   7     9
                                                   3   497           0   500           0   500

J48                    InfoGain/Ranker T = 0.010   9     7   0.643   3    13   0.3     7     9
                                                   3   497           1   499           0   500

J48                    InfoGain/Ranker T = 0.015   9     7   0.643   0    16   0       7     9
                                                   3   497           1   499           0   500

J48                    InfoGain/Ranker T = 0.020   9     7   0.643   0    16   0       7     9
                                                   3   497           1   499           0   500

J48                    InfoGain/Ranker T = 0.025   9     7   0.692   0    16   0       8     8
                                                   1   499           1   499           1   499

J48                    InfoGain/Ranker T = 0.030   9     7   0.692   0    16   0       9     7
                                                   1   499           1   499           1   499

J48                    CfsSubsetEval/BF            9     7   0.72    0    16   0       7     9
                                                   0   500           1   499           0   500
JASPAR CORE     JASPAR CORE + phyloFACTS
         F-mea- Confusion Matrix F-mea-
         sure                    sure



          0.5     6        10   0.414
                  7       493

          0.593   9         7   0.6
                  5       495

          0.593   9         7   0.6
                  5       495

          0.621   10        6   0.667
                  4       496

          0.667   12        4   0.857
                  0       500

          0.692   12        4   0.857
                  0       500

          0.667   11        5   0.815
                  0       500

          0.667   11        5   0.733
                  3       497



          0.476   5        11   0.385
                  5       495

          0.593   5        11   0.385
                  5       495

          0.593   5        11   0.385
                  5       495

          0.593   5        11   0.385
                  5       495

          0.593   5        11   0.385
                  5       495

          0.4     7         9   0.609
                  0       500

          0.615   9         7   0.692
                  1       499
0.615   9      7   0.643
        3    497

0.593   9      7   0.6
        5    495

0.593   9      7   0.6
        5    495

0.4     7      9   0.609
        0    500

0.615   9      7   0.692
        1    499

0.615   9      7   0.643
        3    497

0.593   9      7   0.6
        5    495

0.593   9      7   0.6
        5    495

0.4     7      9   0.609
        0    500

0.64    9      7   0.72
        0    500

0.64    10     6   0.741
        1    499

0.643   10     6   0.69
        3    497

0.643   10     6   0.69
        3    497

0.476   7      9   0.609
        0    500

0.667   10     6   0.769
        0    500

0.667   11     5   0.815
        0    500

0.72    11     5   0.815
        0    500

0.72    11     5   0.815
        0    500

0.476   8      8   0.667
        0    500

0.667   10     6   0.769
        0    500

0.72    11     5   0.815
        0    500

0.72    11     5   0.815
        0    500

0.72    12     4   0.857
        0    500

0.476   7      9   0.609
        0    500

0.667   10     6   0.769
        0    500
0.64    12     4   0.828
        1    499

0.64    10     6   0.714
        2    498

0.64    11     5   0.71
        4    496



0.333   3     13   0.261
        4    496

0.111   1     15   0.118
        0    500

0       0     16   0
        0    500

0.357   5     11   0.435
        2    498

0.357   5     11   0.417
        3    497

0.519   8      8   0.615
        2    498

0.429   8      8   0.615
        2    498

0.643   10     6   0.741
        1    499

0.741   11     5   0.786
        1    499

0.429   5     11   0.417
        3    497



0.207   8      8   0.211
        52   448

0.45    10     6   0.444
        19   481

0.45    10     6   0.444
        19   481

0.545   11     5   0.595
        10   490

0.563   12     4   0.632
        10   490

0.571   11     5   0.564
        12   488

0.6     13     3   0.684
        9    491

0.429   11     5   0.55
        13   487



0.4


0.64    11     5   0.786
        1    499

0.64    11     5   0.786
        1    499

0.667   11     5   0.786
        1    499

0.667   12     4   0.857
        0    500

0.667   12     4   0.857
        0    500

0.667   12     4   0.857
        0    500

0.667   12     4   0.857
        0    500



0.64


0.6     9      7   0.692
        1    499

0.6     9      7   0.667
        2    498

0.571   10     6   0.741
        1    499

0.615   11     5   0.786
        1    499

0.667   11     5   0.815
        0    500

0.667   13     3   0.897
        0    500

0.615




0.48    7      9   0.609
        0    500

0.609   7      9   0.609
        0    500

0.609   7      9   0.609
        0    500

0.609   7      9   0.609
        0    500

0.609   7      9   0.609
        0    500

0.64    8      8   0.667
        0    500

0.692   8      8   0.667
        0    500

0.609

								
To top