categoricalInformationValue

Calculates the value of (P(tag = 1) - P(tag = 0))(log(P(tag = 1)) - log(P(tag = 0))) for each category.

categoricalInformationValue(category1, category2, ..., tag)

The result indicates how a discrete (categorical) feature [category1, category2, ...] contribute to a learning model which predicting the value of tag.