Skip to content

Instantly share code, notes, and snippets.

@abutcher

abutcher/which2 results on UCI datasets

Created April 13, 2010 14:09
Show Gist options
  • Save abutcher/364652 to your computer and use it in GitHub Desktop.
Save abutcher/364652 to your computer and use it in GitHub Desktop.
Download ZIP
Raw
which2 results on UCI datasets
The score being reported is the f-measure. The more number of rounds Which2 runs, the worse the f-measure, but the more stable the results.
Below, it shows that the number of "lives" and the sample size does not matter. It does show that the smaller the beam size, the better the median f-measure.
The median performance, like most classifiers, varies greatly by classifier.
Rounds (Sorted by round)
1 Rounds, 3.0, 10.0, 16.0, 36.0, 100.0, [---- | +++++++++++++++++++++++++++++++++]
2 Rounds, 3.0, 8.0, 14.0, 34.0, 100.0, [--- | ++++++++++++++++++++++++++++++++++]
3 Rounds, 3.0, 7.0, 13.0, 32.0, 100.0, [--- | +++++++++++++++++++++++++++++++++++]
4 Rounds, 3.0, 7.0, 13.0, 30.0, 100.0, [--- | ++++++++++++++++++++++++++++++++++++]
5 Rounds, 3.0, 7.0, 13.0, 32.0, 100.0, [--- | +++++++++++++++++++++++++++++++++++]
6 Rounds, 3.0, 7.0, 13.0, 31.0, 100.0, [--- | +++++++++++++++++++++++++++++++++++ ]
7 Rounds, 3.0, 7.0, 13.0, 31.0, 100.0, [--- | +++++++++++++++++++++++++++++++++++ ]
8 Rounds, 3.0, 7.0, 13.0, 29.0, 100.0, [--- | ++++++++++++++++++++++++++++++++++++ ]
9 Rounds, 3.0, 6.0, 13.0, 31.0, 100.0, [-- | +++++++++++++++++++++++++++++++++++ ]
10 Rounds, 3.0, 6.0, 13.0, 31.0, 100.0, [-- | +++++++++++++++++++++++++++++++++++ ]
11 Rounds, 3.0, 6.0, 13.0, 31.0, 100.0, [-- | +++++++++++++++++++++++++++++++++++ ]
12 Rounds, 3.0, 6.0, 13.0, 31.0, 100.0, [-- | +++++++++++++++++++++++++++++++++++ ]
13 Rounds, 3.0, 6.0, 13.0, 30.0, 100.0, [-- | ++++++++++++++++++++++++++++++++++++]
14 Rounds, 3.0, 6.0, 12.0, 30.0, 100.0, [-- | ++++++++++++++++++++++++++++++++++++]
15 Rounds, 3.0, 6.0, 12.0, 30.0, 100.0, [-- | ++++++++++++++++++++++++++++++++++++]
16 Rounds, 3.0, 6.0, 12.0, 29.0, 100.0, [-- | ++++++++++++++++++++++++++++++++++++ ]
17 Rounds, 3.0, 6.0, 12.0, 30.0, 100.0, [-- | ++++++++++++++++++++++++++++++++++++]
18 Rounds, 3.0, 6.0, 12.0, 29.0, 100.0, [-- | ++++++++++++++++++++++++++++++++++++ ]
19 Rounds, 3.0, 6.0, 12.0, 29.0, 100.0, [-- | ++++++++++++++++++++++++++++++++++++ ]
20 Rounds, 3.0, 6.0, 12.0, 29.0, 100.0, [-- | ++++++++++++++++++++++++++++++++++++ ]
21 Rounds, 3.0, 6.0, 12.0, 29.0, 100.0, [-- | ++++++++++++++++++++++++++++++++++++ ]
22 Rounds, 3.0, 6.0, 12.0, 29.0, 100.0, [-- | ++++++++++++++++++++++++++++++++++++ ]
23 Rounds, 3.0, 6.0, 10.0, 18.0, 53.0, [-- | ++++++++++++++++++ ]
24 Rounds, 3.0, 5.0, 9.0, 15.0, 44.0, [-- | +++++++++++++++ ]
25 Rounds, 3.0, 4.0, 7.0, 14.0, 44.0, [- | ++++++++++++++++ ]
26 Rounds, 3.0, 4.0, 7.0, 13.0, 44.0, [- | ++++++++++++++++ ]
27 Rounds, 3.0, 4.0, 6.0, 12.0, 44.0, [- | +++++++++++++++++ ]
28 Rounds, 3.0, 4.0, 6.0, 11.0, 44.0, [- | +++++++++++++++++ ]
29 Rounds, 3.0, 4.0, 6.0, 11.0, 44.0, [- | +++++++++++++++++ ]
30 Rounds, 3.0, 4.0, 6.0, 10.0, 34.0, [- | +++++++++++++ ]
31 Rounds, 3.0, 4.0, 6.0, 9.0, 30.0, [- |+++++++++++ ]
32 Rounds, 3.0, 7.0, 11.0, 26.0, 61.0, [--- | ++++++++++++++++++ ]
33 Rounds, 3.0, 4.0, 6.0, 9.0, 26.0, [- |+++++++++ ]
34 Rounds, 3.0, 4.0, 5.0, 8.0, 26.0, [-| ++++++++++ ]
35 Rounds, 3.0, 4.0, 5.0, 7.0, 26.0, [-|++++++++++ ]
36 Rounds, 3.0, 4.0, 5.0, 7.0, 26.0, [-|++++++++++ ]
37 Rounds, 3.0, 4.0, 5.0, 7.0, 26.0, [-|++++++++++ ]
38 Rounds, 3.0, 4.0, 5.0, 7.0, 26.0, [-|++++++++++ ]
39 Rounds, 3.0, 4.0, 5.0, 7.0, 26.0, [-|++++++++++ ]
40 Rounds, 3.0, 4.0, 5.0, 6.0, 26.0, [-|+++++++++++ ]
41 Rounds, 3.0, 4.0, 5.0, 6.0, 26.0, [-|+++++++++++ ]
42 Rounds, 3.0, 4.0, 4.0, 6.0, 26.0, [-|+++++++++++ ]
43 Rounds, 3.0, 4.0, 4.0, 6.0, 13.0, [-|++++ ]
44 Rounds, 3.0, 4.0, 4.0, 6.0, 13.0, [-|++++ ]
45 Rounds, 3.0, 4.0, 4.0, 6.0, 13.0, [-|++++ ]
46 Rounds, 3.0, 4.0, 4.0, 6.0, 13.0, [-|++++ ]
47 Rounds, 3.0, 4.0, 4.0, 6.0, 13.0, [-|++++ ]
48 Rounds, 3.0, 4.0, 4.0, 6.0, 13.0, [-|++++ ]
Number of Lives
5, 3.0, 7.0, 13.0, 31.0, 100.0, [--- | +++++++++++++++++++++++++++++++++++ ]
20, 3.0, 7.0, 13.0, 31.0, 100.0, [--- | +++++++++++++++++++++++++++++++++++ ]
15, 3.0, 7.0, 13.0, 31.0, 100.0, [--- | +++++++++++++++++++++++++++++++++++ ]
10, 3.0, 7.0, 13.0, 31.0, 100.0, [--- | +++++++++++++++++++++++++++++++++++ ]
Sample Size
4, 3.0, 7.0, 14.0, 32.0, 100.0, [--- | +++++++++++++++++++++++++++++++++++]
8, 3.0, 7.0, 13.0, 30.0, 100.0, [--- | ++++++++++++++++++++++++++++++++++++]
16, 3.0, 7.0, 13.0, 30.0, 100.0, [--- | ++++++++++++++++++++++++++++++++++++]
32, 3.0, 6.0, 12.0, 30.0, 100.0, [-- | ++++++++++++++++++++++++++++++++++++]
Beam Size
2, 50.0, 51.0, 54.0, 63.0, 100.0, [ - | +++++++++++++++++++ ]
4, 25.0, 25.0, 27.0, 32.0, 100.0, [ -| +++++++++++++++++++++++++++++++++++]
8, 12.0, 12.0, 14.0, 18.0, 100.0, [ -| ++++++++++++++++++++++++++++++++++++++++++]
10, 3.0, 10.0, 11.0, 16.0, 100.0, [----| +++++++++++++++++++++++++++++++++++++++++++]
15, 3.0, 6.0, 8.0, 13.0, 100.0, [-- | ++++++++++++++++++++++++++++++++++++++++++++ ]
20, 3.0, 5.0, 6.0, 12.0, 100.0, [--| +++++++++++++++++++++++++++++++++++++++++++++]
30, 3.0, 3.0, 4.0, 9.0, 100.0, [-| ++++++++++++++++++++++++++++++++++++++++++++++ ]
F-Measure
segment, 16.0, 50.0, 99.0, 100.0, 100.0, [ ------------------ |+]
ionosphere, 6.0, 24.0, 69.0, 100.0, 100.0, [ ---------- | +]
g2, 5.0, 25.0, 40.0, 100.0, 100.0, [ ----------- | +]
vehicle, 6.0, 19.0, 31.0, 44.0, 100.0, [ ------- | +++++++++++++++++++++++++++++]
sonar, 6.0, 20.0, 26.0, 45.0, 75.0, [ -------- | ++++++++++++++++ ]
iris, 5.0, 18.0, 25.0, 39.0, 76.0, [ ------- | +++++++++++++++++++ ]
glass, 3.0, 8.0, 20.0, 32.0, 90.0, [--- | ++++++++++++++++++++++++++++++ ]
pcolic, 7.0, 13.0, 18.0, 32.0, 74.0, [ ---- | ++++++++++++++++++++++ ]
colic, 7.0, 13.0, 18.0, 32.0, 74.0, [ ---- | ++++++++++++++++++++++ ]
labor, 7.0, 12.0, 16.0, 33.0, 55.0, [ --- | ++++++++++++ ]
heart-h, 7.0, 10.0, 16.0, 29.0, 57.0, [ -- | +++++++++++++++ ]
heart-c, 6.0, 11.0, 16.0, 29.0, 56.0, [ --- | ++++++++++++++ ]
splice, 4.0, 10.0, 15.0, 28.0, 76.0, [ ---- | +++++++++++++++++++++++++ ]
breast-cancer, 6.0, 11.0, 15.0, 26.0, 58.0, [ --- | +++++++++++++++++ ]
mushroom, 4.0, 7.0, 13.0, 30.0, 59.0, [ -- | +++++++++++++++ ]
autos, 3.0, 8.0, 13.0, 26.0, 59.0, [--- | +++++++++++++++++ ]
sick, 3.0, 8.0, 12.0, 25.0, 74.0, [--- | +++++++++++++++++++++++++ ]
hypothyroid, 3.0, 6.0, 12.0, 25.0, 74.0, [-- | +++++++++++++++++++++++++ ]
hepatitis, 4.0, 8.0, 12.0, 27.0, 53.0, [ --- | ++++++++++++++ ]
vote, 4.0, 7.0, 11.0, 27.0, 54.0, [ -- | ++++++++++++++ ]
v1, 3.0, 6.0, 11.0, 26.0, 89.0, [-- | ++++++++++++++++++++++++++++++++ ]
lymph, 3.0, 7.0, 11.0, 27.0, 100.0, [--- | +++++++++++++++++++++++++++++++++++++ ]
soybean, 3.0, 5.0, 10.0, 25.0, 64.0, [-- | ++++++++++++++++++++ ]
primary-tumor, 3.0, 6.0, 10.0, 25.0, 100.0, [-- | ++++++++++++++++++++++++++++++++++++++ ]
kr, 3.0, 5.0, 10.0, 25.0, 56.0, [-- | ++++++++++++++++ ]
audiology, 3.0, 5.0, 10.0, 25.0, 100.0, [-- | ++++++++++++++++++++++++++++++++++++++ ]
anneal, 3.0, 5.0, 10.0, 25.0, 64.0, [-- | ++++++++++++++++++++ ]
vowel, 3.0, 5.0, 9.0, 16.0, 100.0, [-- | +++++++++++++++++++++++++++++++++++++++++++]
Overall
3.0, 7.0, 13.0, 31.0,100.0,[--- | +++++++++++++++++++++++++++++++++++ ]
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment