classification: learns classification trees Options: -h : this help message -t filename : specify the training set -p filename : specify the prunning set (absent means no prunning) -T filename : specify the testing set (absent means no testing) -m filename : where to put the resulting tree (default=modelfile) -s filename : where to put the result of testing -M : use multiclassifier instead of normal classifier -D number : number of datasets for multiclassifier -g boolean: do schema matching first -l boolean: labeled choice of split point -d number : minimum no datapoints to consider splittingregression: learns a regression tree Options: -h : this help message -t filename : specify the training set -p filename : specify the prunning set (absent means no prunning) -T filename : specify the testing set (absent means no testing) -m filename : where to put the resulting tree (default=modelfile) -G number : number of isolines in the grid (no effect if testing) -s filename : where to put the result of testing -P : probabilistic regression tree -S number : numer of continuous split variables -r number : number of regressor variables -l number : the biggest nodeID to be used in testing -A : use unidimentional ANOVA splits -L : use multidimentional LDA splits -Q : use multidimentional QDA splits -i number : max number of iterations in EM algorithm -c number : convergence tolerance -R number : number of random restarts for EM -d number : minimum datapoints in leaf
Generated on Mon Jul 21 16:57:23 2003 for SECRET by1.3.2