A Bennett Inequality for the Missing Mass

内容摘要

Novel concentration inequalities are obtained for the missing mass, i.e. the total probability mass of the outcomes not observed in the sample. We derive distribution-free deviation bounds with sublinear exponents in deviation size for missing mass and improve the results of Berend and Kontorovich (2013) and Yari Saeed Khanloo and Haffari (2015) for small deviations which is the most important case in learning theory.