Explainability of AI-Driven Air Combat Agent

Emre Saldiran; Mehmet Hasanzade; Gokhan Inalhan; Antonios Tsourdos

doi:10.1109/CAI54212.2023.00044

Explainability of AI-Driven Air Combat Agent

Emre Saldiran^*, Mehmet Hasanzade, Gokhan Inalhan, Antonios Tsourdos

^*Corresponding author for this work

Cranfield University

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

1 Citation (Scopus)

Abstract

In safety-critical applications, it is crucial to verify and certify the decisions made by AI-driven Autonomous Systems (ASs). However, the black-box nature of neural networks used in these systems often makes it challenging to achieve this. The explainability of these systems can help with the verification and certification process, which will speed up their deployment in safety-critical applications. This study investigates the explainability of AI-driven air combat agents via semantically grouped reward decomposition. The paper presents two use cases to demonstrate how this approach can help AI and non-AI experts to evaluate and debug the behavior of RL agents.

Original language	English
Title of host publication	Proceedings - 2023 IEEE Conference on Artificial Intelligence, CAI 2023
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	85-86
Number of pages	2
ISBN (Electronic)	9798350339840
DOIs	https://doi.org/10.1109/CAI54212.2023.00044
Publication status	Published - 2023
Externally published	Yes
Event	2023 IEEE Conference on Artificial Intelligence, CAI 2023 - Santa Clara, United States Duration: 5 Jun 2023 → 6 Jun 2023

Publication series

Name	Proceedings - 2023 IEEE Conference on Artificial Intelligence, CAI 2023

Conference

Conference	2023 IEEE Conference on Artificial Intelligence, CAI 2023
Country/Territory	United States
City	Santa Clara
Period	5/06/23 → 6/06/23

Bibliographical note

Publisher Copyright:
© 2023 IEEE.

Funding

This work is funded by BAE Systems.

Funders	Funder number
BAE Systems

Keywords

air combat
explainable
reinforcement learning
reward decomposition

Access to Document

10.1109/CAI54212.2023.00044

Cite this

Saldiran, E., Hasanzade, M., Inalhan, G., & Tsourdos, A. (2023). Explainability of AI-Driven Air Combat Agent. In Proceedings - 2023 IEEE Conference on Artificial Intelligence, CAI 2023 (pp. 85-86). (Proceedings - 2023 IEEE Conference on Artificial Intelligence, CAI 2023). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/CAI54212.2023.00044

@inproceedings{6fb3ab424f984b40b4170815030563ca,

title = "Explainability of AI-Driven Air Combat Agent",

abstract = "In safety-critical applications, it is crucial to verify and certify the decisions made by AI-driven Autonomous Systems (ASs). However, the black-box nature of neural networks used in these systems often makes it challenging to achieve this. The explainability of these systems can help with the verification and certification process, which will speed up their deployment in safety-critical applications. This study investigates the explainability of AI-driven air combat agents via semantically grouped reward decomposition. The paper presents two use cases to demonstrate how this approach can help AI and non-AI experts to evaluate and debug the behavior of RL agents.",

keywords = "air combat, explainable, reinforcement learning, reward decomposition",

author = "Emre Saldiran and Mehmet Hasanzade and Gokhan Inalhan and Antonios Tsourdos",

note = "Publisher Copyright: {\textcopyright} 2023 IEEE.; 2023 IEEE Conference on Artificial Intelligence, CAI 2023 ; Conference date: 05-06-2023 Through 06-06-2023",

year = "2023",

doi = "10.1109/CAI54212.2023.00044",

language = "English",

series = "Proceedings - 2023 IEEE Conference on Artificial Intelligence, CAI 2023",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "85--86",

booktitle = "Proceedings - 2023 IEEE Conference on Artificial Intelligence, CAI 2023",

address = "United States",

}

Saldiran, E, Hasanzade, M, Inalhan, G & Tsourdos, A 2023, Explainability of AI-Driven Air Combat Agent. in Proceedings - 2023 IEEE Conference on Artificial Intelligence, CAI 2023. Proceedings - 2023 IEEE Conference on Artificial Intelligence, CAI 2023, Institute of Electrical and Electronics Engineers Inc., pp. 85-86, 2023 IEEE Conference on Artificial Intelligence, CAI 2023, Santa Clara, United States, 5/06/23. https://doi.org/10.1109/CAI54212.2023.00044

Explainability of AI-Driven Air Combat Agent. / Saldiran, Emre; Hasanzade, Mehmet; Inalhan, Gokhan et al.
Proceedings - 2023 IEEE Conference on Artificial Intelligence, CAI 2023. Institute of Electrical and Electronics Engineers Inc., 2023. p. 85-86 (Proceedings - 2023 IEEE Conference on Artificial Intelligence, CAI 2023).

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

TY - GEN

T1 - Explainability of AI-Driven Air Combat Agent

AU - Saldiran, Emre

AU - Hasanzade, Mehmet

AU - Inalhan, Gokhan

AU - Tsourdos, Antonios

PY - 2023

Y1 - 2023

N2 - In safety-critical applications, it is crucial to verify and certify the decisions made by AI-driven Autonomous Systems (ASs). However, the black-box nature of neural networks used in these systems often makes it challenging to achieve this. The explainability of these systems can help with the verification and certification process, which will speed up their deployment in safety-critical applications. This study investigates the explainability of AI-driven air combat agents via semantically grouped reward decomposition. The paper presents two use cases to demonstrate how this approach can help AI and non-AI experts to evaluate and debug the behavior of RL agents.

AB - In safety-critical applications, it is crucial to verify and certify the decisions made by AI-driven Autonomous Systems (ASs). However, the black-box nature of neural networks used in these systems often makes it challenging to achieve this. The explainability of these systems can help with the verification and certification process, which will speed up their deployment in safety-critical applications. This study investigates the explainability of AI-driven air combat agents via semantically grouped reward decomposition. The paper presents two use cases to demonstrate how this approach can help AI and non-AI experts to evaluate and debug the behavior of RL agents.

KW - air combat

KW - explainable

KW - reinforcement learning

KW - reward decomposition

UR - http://www.scopus.com/inward/record.url?scp=85168689179&partnerID=8YFLogxK

U2 - 10.1109/CAI54212.2023.00044

DO - 10.1109/CAI54212.2023.00044

M3 - Conference contribution

AN - SCOPUS:85168689179

T3 - Proceedings - 2023 IEEE Conference on Artificial Intelligence, CAI 2023

SP - 85

EP - 86

BT - Proceedings - 2023 IEEE Conference on Artificial Intelligence, CAI 2023

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2023 IEEE Conference on Artificial Intelligence, CAI 2023

Y2 - 5 June 2023 through 6 June 2023

ER -

Explainability of AI-Driven Air Combat Agent

Abstract

Publication series

Conference

Bibliographical note

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this