An extremal property of rectangular distributions
From MaRDI portal
Bei einer Stichprobe vom Umfang n seien \(x_ i\) and \(x_ j\) die entsprechenden Ordnungsstatistiken. Es wird gezeigt: \[ \max \{corr(f(x_ i),g(x_ j))| \quad f,g\in {\mathcal L}_ 2\}\leq \rho_{i,j} \] mit \(\rho_{i,j}=\{i(n+1-i)/j(n+1-i)\}^{1/2}\). Dazu werden f und g nach Jacobi-Polynomen entwickelt und x (ohne Beschränkung der Allgemeinheit) als gleichverteilt auf (0, 1) angenommen. Für allgemeine stetig verteilte x folgt daraus: \(corr(x_ i,x_ j)\leq \rho_{i,j}\).
Recommendations
- On a characterization of rectangular distributions
- Maximal correlation in a non-diagonal case
- The maximal correlation for the generalized order statistics and dual generalized order statistics
- Upper bound for the covariance of extreme order statistics from a sample of size three
- On Terrell's characterization of uniform distribution
Cited in
(28)- On a characterization of rectangular distributions
- Characterizations of uniform and exponential distributions
- Correlations between functions of records can be negative
- On Terrell's characterization of uniform distribution
- An extremal property of the binomial distribution
- On characterizations based on record values and order statistics
- On a limit relation between order statistics and records
- An extremal property of the normal distribution, with a discrete analog
- Two characterizations based on order statistics and records
- Maximum variance of order statistics
- The maximal correlation for the generalized order statistics and dual generalized order statistics
- Automatic differentiation and maximal correlation of order statistics from discrete parents
- On the correlation structure of exponential order statistics and some extensions
- Some counterexamples concerning maximal correlation and linear regression
- Upper and lower bounds for the correlation ratio of order statistics from a sample without replacement
- Model-free sure screening via maximum correlation
- General bounds and inequalities in order statistics
- A simple method for obtaining the maximal correlation coefficient and related characterizations
- Maximal correlation in a non-diagonal case
- On generalized order statistics and maximal correlation as a measure of dependence
- A note on the upper bound to variance of the sample extreme from a finite population
- On the rate of convergence to asymptotic independence between order statistics under power normalization with extension to the generalized order statistics
- A discrete analogue of Terrell's characterization of rectangular distributions
- Mixed systems with minimal and maximal lifetime variances
- Kendall correlation coefficient for order statistics
- On the maximal correlation of some stochastic processes
- Extreme variances of order statistics in dependent samples
- Characterizations based on order statistics under sampling without replacement
This page was built for publication: An extremal property of rectangular distributions
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1059326)