On the optimality of the simple bayesian classifier under zero-one loss (1997)

by Pedro Domingos, Michael J Pazzani
Venue:Machine Learning