Maximizing expected gain in supervised discrete Bayesian classification when fusing binary valued features.

Robert S. Lynch Jr., Peter K. Willett

Browse the full FUSION paper archive.