Score » Historique » Version 1
François Rioult, 19/10/2013 20:49
1 | 1 | François Rioult | h1. Score |
---|---|---|---|
2 | |||
3 | h2. Synopsis |
||
4 | |||
5 | @score@ computes the validation values from a decision file |
||
6 | |||
7 | h2. Input format |
||
8 | |||
9 | Each lines has three fields, separated by a tab char: |
||
10 | 1. The string description of the object, whose first field is the true class |
||
11 | 2. the decision value for the first class (1) |
||
12 | 3. the decision value for Class 2 |
||
13 | etc. |
||
14 | |||
15 | h2. Example |
||
16 | |||
17 | <pre> |
||
18 | $ cat > /tmp/test |
||
19 | 1 2 |
||
20 | 2 3 |
||
21 | ^D |
||
22 | $ mtminer /tmp/test |
||
23 | 2 |
||
24 | # depth 2 |
||
25 | 1 3 |
||
26 | </pre> |
||
27 | |||
28 | h2. C++ code |
||
29 | |||
30 | The code is avalaible below (mtminer-dist.tgz). |
||
31 | |||
32 | h2. Reference |
||
33 | |||
34 | [1] Hébert C., Bretto A., Crémilleux B. A data mining formalization to improve hypergraph transversal computation, Fundamenta Informaticae, IOS Press, 80(4), pp. 415-433, 2007. |