Projecting statistical functionals (Q5935341)

The present lecture note introduces a projection method for statistical functionals onto convex cones in Hilbert spaces. Sharp bounds in terms of moments of the underlying parameters of the distributions are derived. The eight chapters have the following headlines: 1. Introduction and Notation; 2. Basic Notions; 3. Quantiles; 4. Order Statistics of Independent Samples; 5. Order Statistics of Dependent Observations; 6. Records and \(k\)-th Records; 7. Predictions of Order and Record Statistics; 8. Further Research Directions. The subsequent examples are of particular interest: The upper deviation of the expectation of the \(j\) th statistic from the population mean. Construction of greatest convex minorants. L-statistics. Record values and their occurence times. Conditional expectations of order statistics. Quantile functions, the bias of quantile estimates and hazard failure rates. Dependent observations and bivariate distributions with given marginals. Moreover, different concepts of ordering for distribution functions, densities and failure rates are discussed and compared. The monograph contains a lot of interesting analytic details mostly for order statistics and related quantities.