source: orange/orange/doc/reference/SelfOrganisingMaps.htm @ 6538:a5f65d7f0b2c

Revision 6538:a5f65d7f0b2c, 3.5 KB checked in by Mitar <Mitar@…>, 4 years ago (diff)

Made XPM version of the icon 32x32.

Line 
1<html><HEAD>
2<LINK REL=StyleSheet HREF="../style.css" TYPE="text/css">
3<LINK REL=StyleSheet HREF="style-print.css" TYPE="text/css" MEDIA=print>
4</HEAD> <body>
5
6<h1>Self Organising Maps</h1>
7<index name="classifiers+self organizing maps">
8
9<p>The self-organising map (SOM) is a method for unsupervised learning, based on a grid of artificial neurons whose weights are adapted to match input vectors in a training set.</p>
10
11<h2>SOMLearner</h2>
12<p><INDEX name="classes/SOMLearner">SOMLearner class constructs an instance of <INDEX name="classes/SOMClassifier">SOMClassifier or if given a classless domain an instance of <INDEX name="classes/SOMMap">SOMMap. SOMClassifier is actualy just a SOMMap with a defined call function that returns a majority class at the winning node.</p>
13
14<P class=section>Attributes</P>
15<DL class=attributes>
16<DT>xDim</DT>
17<DD>X dimension of the map (default 10)</DD>
18<DT>yDim</DT>
19<DD>Y dimension of the map (default 10)</DD>
20<DT>topology</DT>
21<DD>Topology of the map. Can be a SOMLearner.RetangularTopology for rectangular or SOMLearner.HexagonalTopology (default) for hexagonal topology</DD>
22<DT>neighborhood<DT>
23<DD>Neighborhood function type. Can be SOMLearner.BubbleNeighborhood (default) or SOMLearner.GaussianNeighborhood</DD>
24<DT>steps</DT>
25<DD>Number of steps (default 2)</DD>
26<DT>alphaType</DT>
27<DD>A alpha function type. Can be a SOMLearner.LinearFunction (default) or SOMLearner.InverseFunction</DD>
28<DT>alpha<DT>
29<DD>A list of alpha values (learning rate) to be used at the beginning of each step (default [0.05,0.03]</DD>
30<DT>iterations</DT>
31<DD>A list of iterations at each step (default [1000, 10000])</DD>
32<DT>radius</DT>
33<DD>A list of radius values for neighborhood function at each step (default [10,5])</DD>
34<DT>domainContinuizer</DT>
35<DD>Domain continuizer used to transform the domain</DD>
36<DT>transformedDomain</DT>
37<DD>Transformed domain</DD>
38<DT>randomSeed</DT>
39<DD>Random seed used to initialize the codebook vectors. Use -1 to use current time as a seed (default 0).</DD>
40</DL>
41
42<h2>SOMMap</h2>
43<p><INDEX name="classes/SOMMap">SOMMap holds the resulting 2 dimensional map of SOMNodes</p>
44<P class=section>Attributes</P>
45<DL class =attributes>
46<DT>xDim</DT>
47<DD>X dimension of the map</DD>
48<DT>yDim</DT>
49<DD>Y dimension of the map</DD>
50<DT>topology</DT>
51<DD>Topology of the map. Can be a SOMLearner.RetangularTopology for rectangular or SOMLearner.HexagonalTopology (default) for hexagonal topology</DD>
52<DT>neighborhood<DT>
53<DD>Neighborhood function type. Can be SOMLearner.BubbleNeighborhood (default) or SOMLearner.GaussianNeighborhood</DD>
54<DT>nodes</DT>
55<DD>A list of SOMNodes</DD>
56<DL>transformedDomain</DL>
57<DD>Transformed domain</DD>
58<DT>error</DT>
59<DD>Quantiztion error of the map</DD>
60</DL>
61
62<P class=section>Methods</P>
63<DL class =attributes>
64<DT>getWinner(example) ((example)->SOMNode)</DT>
65<DD>Returns the node closest to the example</DD>
66</DL>
67
68<h2>SOMNode</h2>
69<p><INDEX name="classes/SOMNode">SOMNode holds the codebook vector</p>
70<P class=section>Attributes</P>
71<DL class =attributes>
72<DT>vector</DT>
73<DD>Holds the codebook vector</DD>
74<DT>examples</DT>
75<DD>Holds the examples for whitch this is the winning node</DD>
76</DT>
77
78<P class =section>Methods</P>
79<DL classs =attributes>
80<DT>getDistance(example) ((example)->float)</DT>
81<DD>Computes the distance to the node</DD>
82</DL>
83
84<h2>Examples</h2>
85<xmp class=code>>>>data=orange.ExampleTable("iris")
86>>>map=orange.SOMLearner(data)
87>>>print map.nodes[0].examples
88...
89...
90>>>print map.nodes[0].vector
91...
92>>>map.getDistance(data[0])
931.56
94</xmp>
95
96</body></html>
97
Note: See TracBrowser for help on using the repository browser.