NAME

heri-split - splits the dataset into training and testing sets

SYNOPSIS

heri-split [OPTIONS] dataset1 [dataset2...]

DESCRIPTION

heri-split splits the dataset into several training and testing sets as it is required for N-fold cross-validation. Dataset contains one object per line as in svmlight format. By default stratified sampling is used. That is, all folds contain the same number of objects for each label. If option -c is specified, testN.txt and trainN.txt files (also in svmlight format) are created, where N is the number of fold. If option -R is specified, test.txt and train.txt files are created for the same purposes. Also testing_fold.txt file is created, where for each object (one per line) its testing fold number is specified if oprion -c is applied. The file testing_fold.txt contain either 1 for testing set and 0 for training set, if option -R is applied.

OPTIONS

-h, --help

Display help information.

-c, --folds count

Set the number of folds. This is a mandatory option.

-d, --output-dir dir

Set the output directory. This is a mandatory option.

-r,--random

Use random sampling instead of stratified one.

-R,--ratio

Split the input dataset into training and testing one in the specified ratio (in percents).

-s, --seed seed

Set the seed value for pseudorandom generator.

HOME

<http://github.com/cheusov/herisvm>

SEE ALSO

heri-eval(1) heri-stat(1)