source: orange/docs/widgets/rst/associate/associationrules.rst @ 11359:8d54e79aa135

Revision 11359:8d54e79aa135, 2.3 KB checked in by Ales Erjavec <ales.erjavec@…>, 14 months ago (diff)

Cleanup of 'Widget catalog' documentation.

Fixed rst text formating, replaced dead hardcoded reference links (now using
:ref:), etc.

Line 
1.. _Association Rules:
2
3Association Rules
4=================
5
6.. image:: ../icons/AssociationRules.png
7
8Construction of Association Rules
9
10Signals
11-------
12
13Inputs:
14   - Examples (ExampleTable)
15      A table with training examples
16
17
18Outputs:
19   - Association Rules
20      Induced rules.
21
22
23
24Description
25-----------
26
27This widget runs several algorithms for induction of association rules:
28the original Agrawal's algorithm for sparse data sets, a modified version
29by Zupan and Demsar which is more suitable for the usual machine learning
30data, and, finally, an algorithm which induces classification rules where
31the right-hand side of the rule is always the class attribute.
32
33.. image:: images/AssociationRules.png
34   :alt: Association Rules Widget
35
36The first check box, :obj:`Induce classification rules` allows you to specify
37the algorithm to use. If checked, the original Agrawal's algorithm is
38used, which is designed for (very) sparse data sets. If clear, it will
39use an algorithm which works better on the usual machine learning data where
40each example is described by a (smaller) list of attributes and there are
41not many missing values.
42
43Next, you can decide whether to :obj:`Induce classification rules` or
44ordinary association rules. The former always have the class attribute
45(and nothing else) on the right-hand side. You can combine this with any
46of the above two algorithms.
47
48As for pruning, you can specify the :obj:`Minimal support` and
49:obj:`Minimal confidence`, where support is percentage of the entire
50data set covered by the (entire) rule and the confidence is the proportion
51of the number of examples which fit the right side among those that fit the
52left side. The running time depends primarily on the support.
53
54If support is set too low, the algorithm may find too many rules and
55eventually run out of memory. For this reason the number of rules is
56by default limited to 10000. You can increase the limit at a risk of
57running out of memory.
58
59:obj:`Build rules` runs the algorithm and outputs the induced rules.
60You need to push this button after changing any settings.
61
62
63Examples
64--------
65
66This widget is typically used with the :ref:`Association Rules Filter`
67and/or :ref:`Association Rules Explorer`. One possible schema is shown below.
68
69.. image:: images/AssociationRules-Schema.png
70   :alt: Association Rules
Note: See TracBrowser for help on using the repository browser.