Class prior change

Performance of several class prior change models was studied across 4 data sets.

Usage

data(prior)

Format

List of 2 data.frames: accuracy contains the mean and standard error of the performance measures (sqErr and accuracy), data.set.info contains meta-data about the dimension and number of positive and negative examples in each data set.

Source

MC du Plessis, personal communication.

References

du Plessis MC, Sugiyama M. Semi-Supervised Learning of Class Balance under Class-Prior Change by Distribution Matching. ICML 2012.