Acyclic discrete phase type distributions: Properties and a parameter estimation algorithm

A. Bobbio, A. Horv́th, M. Scarpa, M. Telek

Research output: Contribution to journalArticle

88 Citations (Scopus)

Abstract

This paper provides a detailed study on discrete phase type (DPH) distributions and its acyclic subclass referred to as acyclic-DPH (ADPH). Previously not considered similarities and differences between DPH and continuous phase type (CPH) distributions are investigated and minimal representations, called canonical forms, for the subclass of ADPH distributions are provided. We investigate the consequences of the recent result about the minimal coefficient of variation of the DPH class [The minimal coefficient of variation of discrete phase type distributions, in: Proceedings of the Third International Conference on Matrix-analytic Methods in Stochastic Models, July 2000] and show that below a given order (that is a function of the expected value) the minimal coefficient of variation of the DPH class is always less than the minimal coefficient of variation of the CPH class. Since all the previously introduced Phase Type fitting methods were designed for fitting over the CPH class we provide a DPH fitting method for the first time. The implementation of the DPH fitting algorithm is found to be simple and stable. The algorithm is tested over a benchmark consisting of 10 different continuous distributions. The error resulted when a continuous distribution sampled in discrete points is fitted by a DPH is also considered.

Original languageEnglish
Pages (from-to)1-32
Number of pages32
JournalPerformance Evaluation
Volume54
Issue number1
DOIs
Publication statusPublished - Sep 2003

Fingerprint

Phase-type Distribution
Discrete Distributions
Estimation Algorithms
Parameter estimation
Parameter Estimation
Stochastic models
Coefficient of variation
Continuous Distributions
Matrix Analytic Methods
Canonical form
Expected Value
Stochastic Model
Benchmark

Keywords

  • Acyclic discrete phase type distributions
  • Canonical forms
  • Continuous and discrete phase type distributions

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Software
  • Modelling and Simulation
  • Statistics and Probability

Cite this

Acyclic discrete phase type distributions : Properties and a parameter estimation algorithm. / Bobbio, A.; Horv́th, A.; Scarpa, M.; Telek, M.

In: Performance Evaluation, Vol. 54, No. 1, 09.2003, p. 1-32.

Research output: Contribution to journalArticle

@article{43a893176f5f4294962fae637d1d09be,
title = "Acyclic discrete phase type distributions: Properties and a parameter estimation algorithm",
abstract = "This paper provides a detailed study on discrete phase type (DPH) distributions and its acyclic subclass referred to as acyclic-DPH (ADPH). Previously not considered similarities and differences between DPH and continuous phase type (CPH) distributions are investigated and minimal representations, called canonical forms, for the subclass of ADPH distributions are provided. We investigate the consequences of the recent result about the minimal coefficient of variation of the DPH class [The minimal coefficient of variation of discrete phase type distributions, in: Proceedings of the Third International Conference on Matrix-analytic Methods in Stochastic Models, July 2000] and show that below a given order (that is a function of the expected value) the minimal coefficient of variation of the DPH class is always less than the minimal coefficient of variation of the CPH class. Since all the previously introduced Phase Type fitting methods were designed for fitting over the CPH class we provide a DPH fitting method for the first time. The implementation of the DPH fitting algorithm is found to be simple and stable. The algorithm is tested over a benchmark consisting of 10 different continuous distributions. The error resulted when a continuous distribution sampled in discrete points is fitted by a DPH is also considered.",
keywords = "Acyclic discrete phase type distributions, Canonical forms, Continuous and discrete phase type distributions",
author = "A. Bobbio and A. Horv́th and M. Scarpa and M. Telek",
year = "2003",
month = "9",
doi = "10.1016/S0166-5316(03)00044-0",
language = "English",
volume = "54",
pages = "1--32",
journal = "Performance Evaluation",
issn = "0166-5316",
publisher = "Elsevier",
number = "1",

}

TY - JOUR

T1 - Acyclic discrete phase type distributions

T2 - Properties and a parameter estimation algorithm

AU - Bobbio, A.

AU - Horv́th, A.

AU - Scarpa, M.

AU - Telek, M.

PY - 2003/9

Y1 - 2003/9

N2 - This paper provides a detailed study on discrete phase type (DPH) distributions and its acyclic subclass referred to as acyclic-DPH (ADPH). Previously not considered similarities and differences between DPH and continuous phase type (CPH) distributions are investigated and minimal representations, called canonical forms, for the subclass of ADPH distributions are provided. We investigate the consequences of the recent result about the minimal coefficient of variation of the DPH class [The minimal coefficient of variation of discrete phase type distributions, in: Proceedings of the Third International Conference on Matrix-analytic Methods in Stochastic Models, July 2000] and show that below a given order (that is a function of the expected value) the minimal coefficient of variation of the DPH class is always less than the minimal coefficient of variation of the CPH class. Since all the previously introduced Phase Type fitting methods were designed for fitting over the CPH class we provide a DPH fitting method for the first time. The implementation of the DPH fitting algorithm is found to be simple and stable. The algorithm is tested over a benchmark consisting of 10 different continuous distributions. The error resulted when a continuous distribution sampled in discrete points is fitted by a DPH is also considered.

AB - This paper provides a detailed study on discrete phase type (DPH) distributions and its acyclic subclass referred to as acyclic-DPH (ADPH). Previously not considered similarities and differences between DPH and continuous phase type (CPH) distributions are investigated and minimal representations, called canonical forms, for the subclass of ADPH distributions are provided. We investigate the consequences of the recent result about the minimal coefficient of variation of the DPH class [The minimal coefficient of variation of discrete phase type distributions, in: Proceedings of the Third International Conference on Matrix-analytic Methods in Stochastic Models, July 2000] and show that below a given order (that is a function of the expected value) the minimal coefficient of variation of the DPH class is always less than the minimal coefficient of variation of the CPH class. Since all the previously introduced Phase Type fitting methods were designed for fitting over the CPH class we provide a DPH fitting method for the first time. The implementation of the DPH fitting algorithm is found to be simple and stable. The algorithm is tested over a benchmark consisting of 10 different continuous distributions. The error resulted when a continuous distribution sampled in discrete points is fitted by a DPH is also considered.

KW - Acyclic discrete phase type distributions

KW - Canonical forms

KW - Continuous and discrete phase type distributions

UR - http://www.scopus.com/inward/record.url?scp=0043123281&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0043123281&partnerID=8YFLogxK

U2 - 10.1016/S0166-5316(03)00044-0

DO - 10.1016/S0166-5316(03)00044-0

M3 - Article

AN - SCOPUS:0043123281

VL - 54

SP - 1

EP - 32

JO - Performance Evaluation

JF - Performance Evaluation

SN - 0166-5316

IS - 1

ER -