Updating Probabilistic Knowledge on Condition/Event Nets using Bayesian Networks
- Submitting institution
-
The University of Leicester
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 1427
- Type
- E - Conference contribution
- DOI
-
10.4230/LIPIcs.CONCUR.2018.27
- Title of conference / published proceedings
- 29th International Conference on Concurrency Theory (CONCUR 2018)
- First page
- 27:1
- Volume
- 118
- Issue
- -
- ISSN
- 1868-8969
- Open access status
- Compliant
- Month of publication
- June
- Year of publication
- 2018
- URL
-
-
- Supplementary information
-
https://doi.org/10.4230/LIPIcs.CONCUR.2018.27
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
3
- Research group(s)
-
-
- Citation count
- -
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- Efficiently representing joint distributions, Bayesian networks are fundamental to reasoning in statistics and machine learning. This paper is the first to formalise the rewriting of BNs, implementing changes over time to the joint distributions. Such manipulation of BNs, informally introduced by Judea Pearl in his Causal Inference approach, models an observer learning about a system while interacting with it. This has applications in cyber-physical systems, security and HCI. The approach is fully developed in the framework of PROPS, symmetric monoidal categories representing both graphical structure semantic distributions of BNS. This guarantees compositionality (Sobocinski et al, CALCO’18) allowing optimisation of evaluation.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -