TREEC

Build a decision tree classifier

    W = TREEC(A,CRIT,PRUNE,T)

Computation of a decision tree classifier out of a dataset A using  a binary splitting criterion CRIT
    INFCRIT - information gain
    MAXCRIT - purity (default)
    FISHCRIT - Fisher criterion

Pruning is defined by prune
    PRUNE = -1 pessimistic pruning as defined by Quinlan.
    PRUNE = -2 testset pruning using the dataset T, or, if not
               supplied, an artificially generated testset of 5 x size of
               the training set based on parzen density estimates.
               see PARZENML and GENDATP.
    PRUNE = 0 no pruning (default).
    PRUNE > 0 early pruning, e.g. prune = 3
    PRUNE = 10 causes heavy pruning.

If CRIT or PRUNE are set to NaN they are optimised by REGOPTC.

See also

datasets, mappings, tree_map, regoptc,