We consider the classical problem of missing-mass estimation, which deals with estimating total probability unseen elements in a sample. The estimation has various applications machine learning, statistics, language processing, ecology, sensor networks, and others. naive, constrained maximum likelihood (CML) estimator is inappropriate for this since it tends to overestimate observed elements. S...

Literature suggests two main approaches, parametric and non-parametric, for constructing efficiency frontiers using which efficiency scores of other units can be based. Parametric functions can be either deterministic or stochastic in nature. However, when multiple inputs and outputs are encountered, Data Envelopment Analysis (DEA), a non-parametric approach, is a powerful tool used for decades...