Abstract
The premise of this paper is that a fractional probability distribution is based on fractional operators and the fractional (Hurst) index used that alters the classical setting of random variables. For example, a random variable defined by its density function might not have a fractional density function defined in its conventional sense. Practically, it implies that a distribution's granularity defined by a fractional kernel may have properties that differ due to the fractional index used and the fractional calculus applied to define it. The purpose of this paper is to consider an application of fractional calculus to define the fractional density function of a random variable. In addition, we provide and prove a number of results, defining the functional forms of these distributions as well as their existence. In particular, we define fractional probability distributions for increasing and decreasing functions that are right continuous. Examples are used to motivate the usefulness of a statistical approach to fractional calculus and its application to economic and financial problems. In conclusion, this paper is a preliminary attempt to construct statistical fractional models. Due to the breadth and the extent of such problems, this paper may be considered as an initial attempt to do so.
Original language | English (US) |
---|---|
Pages (from-to) | 1161-1177 |
Number of pages | 17 |
Journal | Physica A: Statistical Mechanics and its Applications |
Volume | 462 |
DOIs | |
State | Published - Nov 15 2016 |
Keywords
- Finance
- Fractional calculus
- Hurst index
- Modeling
- Randomness
- Statistics
ASJC Scopus subject areas
- Statistics and Probability
- Condensed Matter Physics