Abstract:
Indirect transformations of a one-dimensional, two-dimensional, and multidimensional random variable are proposed. They are based on various symmetrizations of the density function. The focus is on changing peakedness of a distribution about the origin.
Keywords:peakedness, function rearrangement, continuous symmetrization, log-concave density, majorization, statistical inference.
Fulltext:
PDF file (344 kB)