source: orange/Orange/testing/regression/results_modules/statExamples.py.txt @ 9689:ef9a2d79f6e0

Revision 9689:ef9a2d79f6e0, 2.3 KB checked in by Miha Stajdohar <miha.stajdohar@…>, 2 years ago (diff)

New folder structure.

Line 
1
2method  CA  AP  Brier   IS
3bayes   0.903   0.902   0.176    0.758
4tree    0.823   0.822   0.331    0.594
5majrty  0.614   0.526   0.474   -0.000
6
7method  CA  AP  Brier   IS
8bayes   0.903+-0.008    0.902+-0.008    0.176+-0.016     0.758+-0.017
9tree    0.823+-0.015    0.822+-0.016    0.331+-0.033     0.594+-0.033
10majrty  0.614+-0.003    0.526+-0.001    0.474+-0.001    -0.000+-0.000
11
12Confusion matrix for naive Bayes:
13TP: 238, FP: 13, FN: 29.0, TN: 155
14
15Confusion matrix for naive Bayes:
16TP: 240, FP: 18, FN: 27.0, TN: 150
17
18Confusion matrix for naive Bayes for 'van':
19TP: 192, FP: 151, FN: 7.0, TN: 496
20
21Confusion matrix for naive Bayes for 'opel':
22TP: 79, FP: 75, FN: 133.0, TN: 559
23
24    bus van saab    opel
25bus 156 19  17  26
26van 4   192 2   1
27saab    8   68  93  48
28opel    8   64  61  79
29
30Sensitivity and specificity for 'voting'
31method  sens    spec
32bayes   0.891   0.923
33tree    0.798   0.863
34majrty  1.000   0.000
35
36Sensitivity and specificity for 'vehicle=van'
37method  sens    spec
38bayes   0.965   0.767
39tree    0.829   0.964
40majrty  0.000   1.000
41
42AUC (voting)
43     bayes: 0.974
44      tree: 0.925
45    majrty: 0.500
46
47AUC for vehicle using weighted single-out method
48bayes   tree    majority
490.840   0.816   0.500
50
51AUC for vehicle, using different methods
52                            bayes   tree    majority
53       by pairs, weighted:  0.861   0.884   0.500
54                 by pairs:  0.863   0.885   0.500
55    one vs. all, weighted:  0.840   0.816   0.500
56              one vs. all:  0.840   0.816   0.500
57
58AUC for detecting class 'van' in 'vehicle'
590.923   0.897   0.500
60
61AUCs for detecting various classes in 'vehicle'
62bus (218.000) vs others:    0.952   0.941   0.500
63van (199.000) vs others:    0.923   0.897   0.500
64saab (217.000) vs others:   0.737   0.709   0.500
65opel (212.000) vs others:   0.749   0.718   0.500
66
67    bus van saab
68van 0.987
69saab    0.927   0.860
70opel    0.921   0.894   0.587
71
72AUCs for detecting various pairs of classes in 'vehicle'
73van vs bus:     0.987   0.976   0.500
74saab vs bus:    0.927   0.939   0.500
75saab vs van:    0.860   0.904   0.500
76opel vs bus:    0.921   0.958   0.500
77opel vs van:    0.894   0.911   0.500
78opel vs saab:   0.587   0.623   0.500
79
80AUC and SE for voting
81bayes: 0.982+-0.008
82tree: 0.888+-0.025
83majrty: 0.500+-0.045
84
85Difference between naive Bayes and tree: 0.065+-0.066
86
87ROC (first 20 points) for bayes on 'voting'
881.000   1.000
890.970   1.000
900.940   1.000
910.910   1.000
920.896   1.000
930.881   1.000
940.836   1.000
950.821   1.000
960.806   1.000
970.761   1.000
980.746   1.000
990.731   1.000
1000.701   1.000
1010.687   1.000
1020.672   1.000
1030.627   1.000
1040.612   1.000
1050.597   1.000
1060.582   1.000
1070.567   1.000
Note: See TracBrowser for help on using the repository browser.