Series:

Johns Hopkins Studies in the Mathematical Sciences

1 total work

Cover of Nonparametric Probability Density Estimation
The standard methodology of statistics generally assumes that the functional forms of probability densities are known. In practice, such knowledge is seldom available. Consequently, the supposed optimality of classical techniques such as that of maximum likelihood is generally misleading. Even new robust techniques assume strong conditions, such as symmetry and multimodality, for probability densities. In Nonparametric Probability Density Estimation, Richard A. Tapia and James R. Thompson investigate some procedures for estimating densities of unknown functional form, demonstrating how techniques of modern optimization theory can be used for nonparametric density estimation.