iterativeOcc: Train a iterative one-class classifier.

In benmack/iterOneClass: Iterative one-class classification for efficiently exploit unlabeled data

Description

...

Usage

iterativeOcc(train_pos, un, iter_max = 10, n_train_un = nrow(train_pos) * 2,
  k = 10, indep_un = 0.5, expand = 2, folder_out = NULL,
  test_set = NULL, seed = NULL, scale = TRUE, ...)

Arguments

train_pos	a data frame with the positive training samples.
un	an object of class rasterTiled image data to be classified.
iter_max	maximum number of iterations.
n_train_un	the number of unlabeled samples to be used for training and validation.
k	the number of folds used for resampling.
indep_un	the fraction of unlabeled samples used for validation.
expand	...
folder_out	a folder where the results are written to.
test_set	a data frame used as independent test set. the first column must be the response variable with 1 being the positive and -1 the negative samples.
seed	a seed point to be set for sampling the unlabeled data and the
...	other arguments that can be passed to trainOcc

benmack/iterOneClass documentation built on May 12, 2019, 12:57 p.m.