SDRANDFOREST Random forest

  P=SDRANDFOREST(DATA)
  P=SDRANDFOREST(DATA,COUNT)
  P=SDRANDFOREST(DATA,options)

 INPUT
   DATA    Input data set
   COUNT   Number of base classifiers (default: 20)

 OUTPUT
   P       Random forest pipeline

 OPTIONS
  'trees',COUNT  Number of tree classifiers (default: 20)
  'dim'          Number of features selected randomly at each tree stage.
                 dim<1 is used as percentage. (default: 0.2 = 20% of feat.)
  'maxnodes',N   Max.number of nodes used in training (default: 10000)

 FIELDS
   P(1,IND)      Make a subset of base classifiers given classifier indices IND

 DESCRIPTION
 SDRANDFOREST builds a set of COUNT tree classifiers. When constructing a
 base tree classifier, 'dim' randomly-selected features is used at each
 node.

 SEE ALSO
 SDTREE

sdrandforest is referenced in examples:

• kb29: How to choose a classifier for a multi-modal problem
• kb28: Example on building a classifier on database records
• kb22: Note on decision tree performance and speed