http://popups.lib.uliege.be/1373-5411/index.php?id=2463 In the presented study it is shown how heteroscedastic normal variables with unknown variance can be characterized by a symmetric beta distribution of the first kind with a known parameter. The presented variance-free characterization technique is illustrated with testing for normality the empirically observed financial return time series. We further suggest one of the possible extensions of the presented method that can be used for statistical learning with applications in real-time and time-critical systems. Full text issues Volume 19 Operations Research and Risk Management Models fr Thu, 22 Aug 2024 10:17:03 +0200 Thu, 22 Aug 2024 10:17:13 +0200 http://popups.lib.uliege.be/1373-5411/index.php?id=2463 0