Mapping species distributions: Classification, similarity and other methods for presence-only data