The Surprise Index for the Multivariate Normal Distribution
From MaRDI portal
Es sei \(x\) eine Zufallsvariable mit Verteilungsdichte \(p(x)\) und \(x'\) eine entnommene Stichprobe. Dann wird als Überraschungsindex (Surprise Index) \[ \lambda_u= \frac{[E(p(x)^u]^{1/u}}{p(x')},\quad u > 0, \] eingeführt und für mehrdimensionale Normalverteilung diskutiert. Der Überraschungsindex beschreibt die Abweichung einer Stichprobe von dem ,,zu erwartenden Ergebnis bei einer vorgegebenen Hypothese.
Cited in
(9)- The likelihood of various stock market return distributions. I: Principles of inference
- Of bits and wows: a Bayesian theory of surprise with applications to attention
- Bayesian measures of surprise for outlier detection
- A new approach to Bayesian hypothesis testing
- Bayesian ikference procedures derived via the concept of relative surprise
- Analysis of variance for Bayesian inference
- ValidP-Values Behave Exactly as They Should: Some Misleading Criticisms ofP-Values and Their Resolution WithS-Values
- The evidence interval and the Bayesian evidence value: On a unified theory for Bayesian hypothesis testing and interval estimation
- Significance testing with no alternative hypothesis: A measure of surprise
This page was built for publication: The Surprise Index for the Multivariate Normal Distribution
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3234890)
