📄 som_make.html

📁 Kohonen的SOM软件包
💻 HTML
字号:
<!DOCTYPE HTML PUBLIC "-//IETF//DTD HTML//EN"><html><head><title>SOM Toolbox / som_make </title></head><body bgcolor=#f0f0f0><table border=0 width="100%" cellpadding=0 cellspacing=0><tr><td valign=baseline><font size=+2>SOM Toolbox</font></td><td valign=baseline align=center><a href="somtoolbox.html">Online documentation</td><td valign=baseline align=right><a href="http://www.cis.hut.fi/projects/somtoolbox/" target="_top">http://www.cis.hut.fi/projects/somtoolbox/</a></td></tr></table><hr><H1> som_make </H1><H3> Purpose </H3><PRE> Creates, initializes and trains a SOM using default parameters.</PRE><H3> Syntax </H3><UL><PRE>  sMap = som_make(D);  sMap = som_make(...,'argID',value,...);  sMap = som_make(...,value,...);</PRE></UL><H3> Description </H3><PRE> Creates, initializes and trains a SOM with default parameters. Uses functions SOM_TOPOL_STRUCT, SOM_TRAIN_STRUCT, SOM_DATA_STRUCT and SOM_MAP_STRUCT to come up with the default values. First, the number of map units is determined. Unless they are explicitly defined, function SOM_TOPOL_STRUCT is used to determine this. It uses a heuristic formula of 'munits = 5*dlen^0.54321'. The 'mapsize' argument influences the final number of map units: a 'big' map has  x4 the default number of map units and a 'small' map has x0.25 the default number of map units.  After the number of map units has been determined, the map size is  determined. Basically, the two biggest eigenvalues of the training data are calculated and the ratio between sidelengths of the map grid is set to this ratio. The actual sidelengths are then set so that  their product is as close to the desired number of map units as possible. Then the SOM is initialized. First, linear initialization along two greatest eigenvectors is tried, but if this can't be done (the eigenvectors cannot be calculated), random initialization is used instead.  After initialization, the SOM is trained in two phases: first rough training and then fine-tuning. If the 'tracking' argument is greater than zero, the average quantization error and topographic error of the final map are calculated.</PRE><H3> Required input arguments </H3><PRE>  D           The data to use in the training.     (struct) A data struct. If a struct is given, '.comp_names' field as               well as '.comp_norm' field is copied to the map struct.     (matrix) A data matrix, size dlen x dim. The data matrix may              contain unknown values, indicated by NaNs. </PRE><H3> Optional input arguments </H3><PRE>  argID (string) Argument identifier string (see below).  value (varies) Value for the argument (see below). Here are the valid argument IDs and corresponding values. The values  which are unambiguous (marked with '*') can be given without the preceeding argID.   'init'       *(string) initialization: 'randinit' or 'lininit' (default)   'algorithm'  *(string) training: 'seq' or 'batch' (default) or 'sompak'   'munits'      (scalar) the preferred number of map units   'msize'       (vector) map grid size   'mapsize'    *(string) do you want a 'small', 'normal' or 'big' map                          Any explicit settings of munits or msize override this.   'lattice'    *(string) map lattice, 'hexa' or 'rect'   'shape'      *(string) map shape, 'sheet', 'cyl' or 'toroid'   'neigh'      *(string) neighborhood function, 'gaussian', 'cutgauss',                          'ep' or 'bubble'   'topol'      *(struct) topology struct   'som_topol','sTopol' = 'topol'   'mask'        (vector) BMU search mask, size dim x 1   'name'        (string) map name   'comp_names'  (string array / cellstr) component names, size dim x 1   'tracking'    (scalar) how much to report, default = 1   'training'    (string) 'short', 'default' or 'long'                 (vector) size 1 x 2, first length of rough training in epochs,                           and then length of finetuning in epochs</PRE><H3> Output arguments </H3><PRE>  sMap (struct) the trained map struct</PRE><H3> Examples </H3><PRE>  To simply train a map with default parameters:    sMap = som_make(D);   With the optional arguments, the initialization and training can be  influenced. To change map size, use 'msize', 'munits' or 'mapsize'  arguments:     sMap = som_make(D,'mapsize','big'); or sMap=som_make(D,'big');   sMap = som_make(D,'munits', 100);   sMap = som_make(D,'msize', [20 10]);   Argument 'algorithm' can be used to switch between 'seq' and 'batch'  algorithms. 'batch' is the default, so to use 'seq' algorithm:    sMap = som_make(D,'algorithm','seq'); or sMap = som_make(D,'seq');   The 'tracking' argument can be used to control the amout of reporting  during training. The argument is used in this function, and it is  passed to the training functions. To make the function work silently  set it to 0.   sMap = som_make(D,'tracking',0); </PRE><H3> See also </H3><TABLE NOBORDER WIDTH=80%><TR><TD><a href="som_map_struct.html"><B>som_map_struct</B></a><TD> Create a map struct.<TR><TD><a href="som_topol_struct.html"><B>som_topol_struct</B></a><TD> Default values for SOM topology.<TR><TD><a href="som_train_struct.html"><B>som_train_struct</B></a><TD> Default values for SOM training parameters.<TR><TD><a href="som_randinint.html"><B>som_randinint</B></a><TD> Random initialization algorithm.<TR><TD><a href="som_lininit.html"><B>som_lininit</B></a><TD> Linear initialization algorithm.<TR><TD><a href="som_seqtrain.html"><B>som_seqtrain</B></a><TD> Sequential training algorithm.<TR><TD><a href="som_batchtrain.html"><B>som_batchtrain</B></a><TD> Batch training algorithm.</TABLE><p><hr><br><br><!-- Last updated: May 30 2002 --></body></html>
⌨️ 快捷键说明

复制代码 Ctrl + C
搜索代码 Ctrl + F
全屏模式 F11
切换主题 Ctrl + Shift + D
显示快捷键 ?
增大字号 Ctrl + =
减小字号 Ctrl + -