lpfgarcia/born: Generate Borderline Noise for Classification Problems

Provides new methods for the artificial addition of label noise in the borderline examples to add more challenging and yet realistic artificial noise to classification datasets. The main difference between the methods is the criterion and bias adopted to estimate which are the borderline examples to be disturbed: the first method is based on the ratio of intra/inter class Nearest Neighbor distance and the second method is based on the distance between the examples and the decision border induced by a radial kernel.

install.packages("remotes")
remotes::install_github("lpfgarcia/born")