Monte-Carlo approximations for Dempster-Shafer belief theoretic algorithms

Thanuka L. Wickramarathne, Kamal Premaratne, Manohar Murthi

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

6 Citations (Scopus)

Abstract

The Dempster-Shafer (DS) belief theory is often used within data fusion, particularly in applications rife with uncertainty that causes problems for probabilistic models. However, when a large number of variables is involved, DS theory (DST) based techniques can quickly become intractable. In this paper, we present a method for complexity reduction of DST methods based on statistical sampling, a tool commonly used in probabilistic-based signal processing (e.g., particle filters). In particular, we use sampling-based approximations to reduce the number of propositions with non-zero support, upon which the computational complexity of many DST-based algorithms are directly dependent on, thereby significantly reducing the computational overhead. We present some preliminary results that demonstrate the validity and accuracy of the proposed method, along with some insights into further developments. We also compare the proposed method to previously presented approximation methods.

Original languageEnglish
Title of host publicationFusion 2011 - 14th International Conference on Information Fusion
StatePublished - Sep 13 2011
Event14th International Conference on Information Fusion, Fusion 2011 - Chicago, IL, United States
Duration: Jul 5 2011Jul 8 2011

Other

Other14th International Conference on Information Fusion, Fusion 2011
CountryUnited States
CityChicago, IL
Period7/5/117/8/11

Fingerprint

Sampling
Data fusion
Computational complexity
Signal processing
Uncertainty
Statistical Models

Keywords

  • Computational complexity
  • Core approximation
  • Dempster-Shafer Theory (DST)
  • Importance sampling

ASJC Scopus subject areas

  • Information Systems

Cite this

Wickramarathne, T. L., Premaratne, K., & Murthi, M. (2011). Monte-Carlo approximations for Dempster-Shafer belief theoretic algorithms. In Fusion 2011 - 14th International Conference on Information Fusion [5977678]

Monte-Carlo approximations for Dempster-Shafer belief theoretic algorithms. / Wickramarathne, Thanuka L.; Premaratne, Kamal; Murthi, Manohar.

Fusion 2011 - 14th International Conference on Information Fusion. 2011. 5977678.

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

Wickramarathne, TL, Premaratne, K & Murthi, M 2011, Monte-Carlo approximations for Dempster-Shafer belief theoretic algorithms. in Fusion 2011 - 14th International Conference on Information Fusion., 5977678, 14th International Conference on Information Fusion, Fusion 2011, Chicago, IL, United States, 7/5/11.
Wickramarathne TL, Premaratne K, Murthi M. Monte-Carlo approximations for Dempster-Shafer belief theoretic algorithms. In Fusion 2011 - 14th International Conference on Information Fusion. 2011. 5977678
Wickramarathne, Thanuka L. ; Premaratne, Kamal ; Murthi, Manohar. / Monte-Carlo approximations for Dempster-Shafer belief theoretic algorithms. Fusion 2011 - 14th International Conference on Information Fusion. 2011.
@inproceedings{54694ea7167a421b90c94c6e54e55d77,
title = "Monte-Carlo approximations for Dempster-Shafer belief theoretic algorithms",
abstract = "The Dempster-Shafer (DS) belief theory is often used within data fusion, particularly in applications rife with uncertainty that causes problems for probabilistic models. However, when a large number of variables is involved, DS theory (DST) based techniques can quickly become intractable. In this paper, we present a method for complexity reduction of DST methods based on statistical sampling, a tool commonly used in probabilistic-based signal processing (e.g., particle filters). In particular, we use sampling-based approximations to reduce the number of propositions with non-zero support, upon which the computational complexity of many DST-based algorithms are directly dependent on, thereby significantly reducing the computational overhead. We present some preliminary results that demonstrate the validity and accuracy of the proposed method, along with some insights into further developments. We also compare the proposed method to previously presented approximation methods.",
keywords = "Computational complexity, Core approximation, Dempster-Shafer Theory (DST), Importance sampling",
author = "Wickramarathne, {Thanuka L.} and Kamal Premaratne and Manohar Murthi",
year = "2011",
month = "9",
day = "13",
language = "English",
isbn = "9781457702679",
booktitle = "Fusion 2011 - 14th International Conference on Information Fusion",

}

TY - GEN

T1 - Monte-Carlo approximations for Dempster-Shafer belief theoretic algorithms

AU - Wickramarathne, Thanuka L.

AU - Premaratne, Kamal

AU - Murthi, Manohar

PY - 2011/9/13

Y1 - 2011/9/13

N2 - The Dempster-Shafer (DS) belief theory is often used within data fusion, particularly in applications rife with uncertainty that causes problems for probabilistic models. However, when a large number of variables is involved, DS theory (DST) based techniques can quickly become intractable. In this paper, we present a method for complexity reduction of DST methods based on statistical sampling, a tool commonly used in probabilistic-based signal processing (e.g., particle filters). In particular, we use sampling-based approximations to reduce the number of propositions with non-zero support, upon which the computational complexity of many DST-based algorithms are directly dependent on, thereby significantly reducing the computational overhead. We present some preliminary results that demonstrate the validity and accuracy of the proposed method, along with some insights into further developments. We also compare the proposed method to previously presented approximation methods.

AB - The Dempster-Shafer (DS) belief theory is often used within data fusion, particularly in applications rife with uncertainty that causes problems for probabilistic models. However, when a large number of variables is involved, DS theory (DST) based techniques can quickly become intractable. In this paper, we present a method for complexity reduction of DST methods based on statistical sampling, a tool commonly used in probabilistic-based signal processing (e.g., particle filters). In particular, we use sampling-based approximations to reduce the number of propositions with non-zero support, upon which the computational complexity of many DST-based algorithms are directly dependent on, thereby significantly reducing the computational overhead. We present some preliminary results that demonstrate the validity and accuracy of the proposed method, along with some insights into further developments. We also compare the proposed method to previously presented approximation methods.

KW - Computational complexity

KW - Core approximation

KW - Dempster-Shafer Theory (DST)

KW - Importance sampling

UR - http://www.scopus.com/inward/record.url?scp=80052548574&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=80052548574&partnerID=8YFLogxK

M3 - Conference contribution

SN - 9781457702679

BT - Fusion 2011 - 14th International Conference on Information Fusion

ER -