Abstract
The class of distributions on [0, ∞) having a rational Laplace transform (i.e., a Laplace transform that is the fraction between two polynomials) will, for reasons that will become apparent in the next chapter, be referred to as matrix-exponential distributions. Within the class of matrix-exponential distributions there is the subclass of phase-type distributions, which are defined in terms of an underlying Markov jump process (or Markov chain). As opposed to a general matrix-exponential distribution, we can for a phase-type distribution use the behavior of the underlying Markov jump process (or chain) in the deduction of its properties and in applications. Deduction in which we condition on an underlying Markov process is often referred to as probabilistic reasoning, as opposed to deduction in the general class of matrix-exponential distributions, where more analytic methods may be necessary. The probabilistic reasoning provides both elegance and power to the theory of matrix-exponential methods and to applications in stochastic modeling.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
We could here have chosen to continue with n-tuples, letting one of the coordinates be equal to the absorbing state of the absorbed process. However, since the absorbed process is no longer needed, we prefer to make a reduction of the state space.
- 2.
see [103], p. 270, for details.
- 3.
For the following calculations, Y may in fact be any distribution with a lighter tail than N.
References
Higham, N.J.: Functions of matrices. Society for Industrial and Applied Mathematics (SIAM), Philadelphia, PA (2008)
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 2017 Springer Science+Business Media LLC
About this chapter
Cite this chapter
Bladt, M., Nielsen, B.F. (2017). Phase-Type Distributions. In: Matrix-Exponential Distributions in Applied Probability. Probability Theory and Stochastic Modelling, vol 81. Springer, Boston, MA. https://doi.org/10.1007/978-1-4939-7049-0_3
Download citation
DOI: https://doi.org/10.1007/978-1-4939-7049-0_3
Published:
Publisher Name: Springer, Boston, MA
Print ISBN: 978-1-4939-7047-6
Online ISBN: 978-1-4939-7049-0
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)