Understanding Heavy Tails in a Bounded World Or, Is a Truncated Heavy Tail Heavy or Not?
نویسندگان
چکیده
We address the important question of the extent to which random variables and vectors with truncated power tails retain the characteristic features of random variables and vectors with power tails. We define two truncation regimes, soft truncation regime and hard truncation regime, and show that, in the soft truncation regime, truncated power tails behave, in important respects, as if no truncation took place. On the other hand, in the hard truncation regime much of “heavy tailedness” is lost. We show how to estimate consistently the tail exponent when the tails are truncated, and suggest statistical tests to decide on whether the truncation is soft or hard. Finally, we apply our methods to two recent data sets arising from computer networks.
منابع مشابه
Variable Heavy Tailed Durations in Internet Traffic, Part I: Understanding Heavy Tails
This paper is part of a larger paper that studies tails of the duration distribution of Internet data flows, and their “heaviness”. Data analysis motivates the concepts of moderate, far and extreme tails for understanding the richness of information available in the data. The analysis also motivates a notion of “variable tail index”, which leads to a generalization of existing theory for heavy ...
متن کاملDo Financial Returns Have Finite or Infinite Variance? a Paradox and an Explanation
One of the major points of contention in studying and modeling nancial returns is whether or not the variance of the returns is nite or in nite (sometimes referred to as the Bachelier-Samuelson Gaussian world versus the Mandelbrot stable world). A di erent formulation of the question asks how heavy the tails of the nancial returns are. The available empirical evidence can be, and has been, inte...
متن کاملSelf-similar Communication Models and Very Heavy Tails
Several studies of le sizes either being downloaded or stored in the world wide web have commented that tails can be so heavy that not only are variances innnite, but so are means. Motivated by this fact, we study the innnite node Poisson model under the assumption that transmission times are heavy tailed with innnite mean. The model is unstable but we are able to provide growth rates. Self-sim...
متن کاملHeavy Traffic Methods in Wireless Systems: towards Modeling Heavy Tails and Long Range Dependence
Heavy traffic models for wireless queueing systems under short range dependence and light tail assumptions on the data traffic have been studied recently. We outline one such model considered by Buche and Kushner [7]. At the same time, similarly to what happened for wireline networks, the emergence of high capacity applications (multimedia, gaming) and inherent mechanisms (multi-access interfer...
متن کاملVariable Heavy Tailed Durations in Internet Traffic, Part II: Theoretical Implications
This paper is part of a larger paper that studies tails of the duration distribution of Internet data flows, and their “heaviness”. Data analysis motivates the concepts of moderate, far and extreme tails for understanding the richness of information available in the data. The data analysis also motivates a notion of “variable tail index”, which leads to a generalization of existing theory for h...
متن کامل