Using observation ageing to improve Markovian model learning in QoS engineering

Radu Calinescu*, Kenneth Johnson, Yasmin Rafiq

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Markovian models are widely used to analyse quality-of-service properties of both system designs and deployed systems. Thanks to the emergence of probabilistic model checkers, this analysis can be performed with high accuracy. However, its usefulness is heavily dependent on how well the model captures the actual behaviour of the analysed system. Our work addresses this problem for a class of Markovian models termed discrete-time Markov chains (DTMCs). We propose a new Bayesian technique for learning the state transition probabilities of DTMCs based on observations of the modelled system. Unlike existing approaches, our technique weighs observations based on their age, to account for the fact that older observations are less relevant than more recent ones. A case study from the area of bioinformatics workflows demonstrates the effectiveness of the technique in scenarios where the model parameters change over time.

Original languageEnglish
Title of host publicationICPE'11 - Proceedings of the 2nd Joint WOSP/SIPEW International Conference on Performance Engineering
Place of PublicationNew York, NY (US)
PublisherACM
Pages505-510
Number of pages6
ISBN (Print)978-1-4503-0519-8
DOIs
Publication statusPublished - 14 Mar 2011
Event2nd Joint WOSP/SIPEW International Conference on Performance Engineering - Karlsruhe, Germany
Duration: 14 Mar 201116 Mar 2011

Conference

Conference2nd Joint WOSP/SIPEW International Conference on Performance Engineering
Abbreviated titleICPE 2011
CountryGermany
CityKarlsruhe
Period14/03/1116/03/11

Fingerprint

Quality of service
Aging of materials
Markov processes
Bioinformatics
Systems analysis

Keywords

  • Algorithms
  • Measurement
  • Reliability
  • Theory

Cite this

Calinescu, R., Johnson, K., & Rafiq, Y. (2011). Using observation ageing to improve Markovian model learning in QoS engineering. In ICPE'11 - Proceedings of the 2nd Joint WOSP/SIPEW International Conference on Performance Engineering (pp. 505-510). New York, NY (US): ACM. https://doi.org/10.1145/1958746.1958823
Calinescu, Radu ; Johnson, Kenneth ; Rafiq, Yasmin. / Using observation ageing to improve Markovian model learning in QoS engineering. ICPE'11 - Proceedings of the 2nd Joint WOSP/SIPEW International Conference on Performance Engineering. New York, NY (US) : ACM, 2011. pp. 505-510
@inproceedings{99e63566aabc46b9bfbf35aa0f199974,
title = "Using observation ageing to improve Markovian model learning in QoS engineering",
abstract = "Markovian models are widely used to analyse quality-of-service properties of both system designs and deployed systems. Thanks to the emergence of probabilistic model checkers, this analysis can be performed with high accuracy. However, its usefulness is heavily dependent on how well the model captures the actual behaviour of the analysed system. Our work addresses this problem for a class of Markovian models termed discrete-time Markov chains (DTMCs). We propose a new Bayesian technique for learning the state transition probabilities of DTMCs based on observations of the modelled system. Unlike existing approaches, our technique weighs observations based on their age, to account for the fact that older observations are less relevant than more recent ones. A case study from the area of bioinformatics workflows demonstrates the effectiveness of the technique in scenarios where the model parameters change over time.",
keywords = "Algorithms, Measurement, Reliability, Theory",
author = "Radu Calinescu and Kenneth Johnson and Yasmin Rafiq",
year = "2011",
month = "3",
day = "14",
doi = "10.1145/1958746.1958823",
language = "English",
isbn = "978-1-4503-0519-8",
pages = "505--510",
booktitle = "ICPE'11 - Proceedings of the 2nd Joint WOSP/SIPEW International Conference on Performance Engineering",
publisher = "ACM",
address = "United States",

}

Calinescu, R, Johnson, K & Rafiq, Y 2011, Using observation ageing to improve Markovian model learning in QoS engineering. in ICPE'11 - Proceedings of the 2nd Joint WOSP/SIPEW International Conference on Performance Engineering. ACM, New York, NY (US), pp. 505-510, 2nd Joint WOSP/SIPEW International Conference on Performance Engineering, Karlsruhe, Germany, 14/03/11. https://doi.org/10.1145/1958746.1958823

Using observation ageing to improve Markovian model learning in QoS engineering. / Calinescu, Radu; Johnson, Kenneth; Rafiq, Yasmin.

ICPE'11 - Proceedings of the 2nd Joint WOSP/SIPEW International Conference on Performance Engineering. New York, NY (US) : ACM, 2011. p. 505-510.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Using observation ageing to improve Markovian model learning in QoS engineering

AU - Calinescu, Radu

AU - Johnson, Kenneth

AU - Rafiq, Yasmin

PY - 2011/3/14

Y1 - 2011/3/14

N2 - Markovian models are widely used to analyse quality-of-service properties of both system designs and deployed systems. Thanks to the emergence of probabilistic model checkers, this analysis can be performed with high accuracy. However, its usefulness is heavily dependent on how well the model captures the actual behaviour of the analysed system. Our work addresses this problem for a class of Markovian models termed discrete-time Markov chains (DTMCs). We propose a new Bayesian technique for learning the state transition probabilities of DTMCs based on observations of the modelled system. Unlike existing approaches, our technique weighs observations based on their age, to account for the fact that older observations are less relevant than more recent ones. A case study from the area of bioinformatics workflows demonstrates the effectiveness of the technique in scenarios where the model parameters change over time.

AB - Markovian models are widely used to analyse quality-of-service properties of both system designs and deployed systems. Thanks to the emergence of probabilistic model checkers, this analysis can be performed with high accuracy. However, its usefulness is heavily dependent on how well the model captures the actual behaviour of the analysed system. Our work addresses this problem for a class of Markovian models termed discrete-time Markov chains (DTMCs). We propose a new Bayesian technique for learning the state transition probabilities of DTMCs based on observations of the modelled system. Unlike existing approaches, our technique weighs observations based on their age, to account for the fact that older observations are less relevant than more recent ones. A case study from the area of bioinformatics workflows demonstrates the effectiveness of the technique in scenarios where the model parameters change over time.

KW - Algorithms

KW - Measurement

KW - Reliability

KW - Theory

UR - http://www.scopus.com/inward/record.url?scp=79953842360&partnerID=8YFLogxK

U2 - 10.1145/1958746.1958823

DO - 10.1145/1958746.1958823

M3 - Conference contribution

AN - SCOPUS:79953842360

SN - 978-1-4503-0519-8

SP - 505

EP - 510

BT - ICPE'11 - Proceedings of the 2nd Joint WOSP/SIPEW International Conference on Performance Engineering

PB - ACM

CY - New York, NY (US)

ER -

Calinescu R, Johnson K, Rafiq Y. Using observation ageing to improve Markovian model learning in QoS engineering. In ICPE'11 - Proceedings of the 2nd Joint WOSP/SIPEW International Conference on Performance Engineering. New York, NY (US): ACM. 2011. p. 505-510 https://doi.org/10.1145/1958746.1958823