Bayesian Information Fusion for Imprecise Probabilistic Models with Different Types of Information

Chenxing Wang; Lechang Yang; R. Rocchetta

doi:10.1109/SRSE54209.2021.00066

Bayesian Information Fusion for Imprecise Probabilistic Models with Different Types of Information

Chenxing Wang
, Lechang Yang
, R. Rocchetta

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

Abstract

A novel approximate Bayesian information fusion method is proposed based on Wasserstein distance and it is applied it to imprecise probabilistic models with different types of information. The proposed method combines the principle of maximum relative entropy with approximate Bayesian computation and uses the Wasserstein distance to perform the approximate Bayesian computations. The key benefit of this approach is the capability to handle different types of information, such as point observed data and moment information. To verify the effectiveness of the proposed method, we apply it to a simple supported beam problem. The results are analyzed towards accuracy and the proposed method is compared to the classical Bayesian approach combined with maximum relative entropy.

Original language	English
Title of host publication	2021 3rd International Conference on System Reliability and Safety Engineering (SRSE)
Publisher	Institute of Electrical and Electronics Engineers
Pages	363-368
Number of pages	6
ISBN (Electronic)	978-1-6654-0160-9
DOIs	https://doi.org/10.1109/SRSE54209.2021.00066
Publication status	Published - 4 Feb 2022
Event	2021 3rd International Conference on System Reliability and Safety Engineering (SRSE) - Harbin, China Duration: 26 Nov 2021 → 28 Nov 2021

Conference

Conference	2021 3rd International Conference on System Reliability and Safety Engineering (SRSE)
Country/Territory	China
City	Harbin
Period	26/11/21 → 28/11/21

Keywords

Knowledge engineering
Fuses
Probabilistic logic
Entropy
Bayes methods
Calibration
Safety
imprecise probability
Maximum relative entropy
Approximate Bayesian computation
Information fusion
Wasserstein distance

Access to Document

10.1109/SRSE54209.2021.00066

Cite this

@inproceedings{7040c1b4046a4b8390843c16ae6adb5d,

title = "Bayesian Information Fusion for Imprecise Probabilistic Models with Different Types of Information",

abstract = "A novel approximate Bayesian information fusion method is proposed based on Wasserstein distance and it is applied it to imprecise probabilistic models with different types of information. The proposed method combines the principle of maximum relative entropy with approximate Bayesian computation and uses the Wasserstein distance to perform the approximate Bayesian computations. The key benefit of this approach is the capability to handle different types of information, such as point observed data and moment information. To verify the effectiveness of the proposed method, we apply it to a simple supported beam problem. The results are analyzed towards accuracy and the proposed method is compared to the classical Bayesian approach combined with maximum relative entropy.",

keywords = "Knowledge engineering, Fuses, Probabilistic logic, Entropy, Bayes methods, Calibration, Safety, imprecise probability, Maximum relative entropy, Approximate Bayesian computation, Information fusion, Wasserstein distance",

author = "Chenxing Wang and Lechang Yang and R. Rocchetta",

year = "2022",

month = feb,

day = "4",

doi = "10.1109/SRSE54209.2021.00066",

language = "English",

pages = "363--368",

booktitle = "2021 3rd International Conference on System Reliability and Safety Engineering (SRSE)",

publisher = "Institute of Electrical and Electronics Engineers",

address = "United States",

note = "2021 3rd International Conference on System Reliability and Safety Engineering (SRSE) ; Conference date: 26-11-2021 Through 28-11-2021",

}

Wang, C, Yang, L & Rocchetta, R 2022, Bayesian Information Fusion for Imprecise Probabilistic Models with Different Types of Information. in 2021 3rd International Conference on System Reliability and Safety Engineering (SRSE)., 9694395, Institute of Electrical and Electronics Engineers, pp. 363-368, 2021 3rd International Conference on System Reliability and Safety Engineering (SRSE), Harbin, China, 26/11/21. https://doi.org/10.1109/SRSE54209.2021.00066

Bayesian Information Fusion for Imprecise Probabilistic Models with Different Types of Information. / Wang, Chenxing; Yang, Lechang; Rocchetta, R.
2021 3rd International Conference on System Reliability and Safety Engineering (SRSE). Institute of Electrical and Electronics Engineers, 2022. p. 363-368 9694395.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - Bayesian Information Fusion for Imprecise Probabilistic Models with Different Types of Information

AU - Wang, Chenxing

AU - Yang, Lechang

AU - Rocchetta, R.

PY - 2022/2/4

Y1 - 2022/2/4

N2 - A novel approximate Bayesian information fusion method is proposed based on Wasserstein distance and it is applied it to imprecise probabilistic models with different types of information. The proposed method combines the principle of maximum relative entropy with approximate Bayesian computation and uses the Wasserstein distance to perform the approximate Bayesian computations. The key benefit of this approach is the capability to handle different types of information, such as point observed data and moment information. To verify the effectiveness of the proposed method, we apply it to a simple supported beam problem. The results are analyzed towards accuracy and the proposed method is compared to the classical Bayesian approach combined with maximum relative entropy.

AB - A novel approximate Bayesian information fusion method is proposed based on Wasserstein distance and it is applied it to imprecise probabilistic models with different types of information. The proposed method combines the principle of maximum relative entropy with approximate Bayesian computation and uses the Wasserstein distance to perform the approximate Bayesian computations. The key benefit of this approach is the capability to handle different types of information, such as point observed data and moment information. To verify the effectiveness of the proposed method, we apply it to a simple supported beam problem. The results are analyzed towards accuracy and the proposed method is compared to the classical Bayesian approach combined with maximum relative entropy.

KW - Knowledge engineering

KW - Fuses

KW - Probabilistic logic

KW - Entropy

KW - Bayes methods

KW - Calibration

KW - Safety

KW - imprecise probability

KW - Maximum relative entropy

KW - Approximate Bayesian computation

KW - Information fusion

KW - Wasserstein distance

UR - https://www.scopus.com/pages/publications/85126228906

U2 - 10.1109/SRSE54209.2021.00066

DO - 10.1109/SRSE54209.2021.00066

M3 - Conference contribution

SP - 363

EP - 368

BT - 2021 3rd International Conference on System Reliability and Safety Engineering (SRSE)

PB - Institute of Electrical and Electronics Engineers

T2 - 2021 3rd International Conference on System Reliability and Safety Engineering (SRSE)

Y2 - 26 November 2021 through 28 November 2021

ER -

Bayesian Information Fusion for Imprecise Probabilistic Models with Different Types of Information

Abstract

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this