Explaining Control Policies through Predicate Decision Diagrams

D 2025

Explaining Control Policies through Predicate Decision Diagrams

CHAKRABORTY, Debraj; Clemens DUBSLAFF; Sudeep KANAV; Jan KŘETÍNSKÝ; Christoph WEINHUBER et al.

Základní údaje

Originální název

Explaining Control Policies through Predicate Decision Diagrams

Název česky

Explaining Control Policies through Predicate Decision Diagrams

Autoři

CHAKRABORTY, Debraj; Clemens DUBSLAFF; Sudeep KANAV; Jan KŘETÍNSKÝ a Christoph WEINHUBER

Vydání

Irvine, USA, HSCC '25: Proceedings of the 28th ACM International Conference on Hybrid Systems: Computation and Control, od s. 1-12, 12 s. 2025

Nakladatel

Association for Computing Machinery

Další údaje

Jazyk

angličtina

Typ výsledku

Stať ve sborníku

Obor

10201 Computer sciences, information science, bioinformatics

Utajení

není předmětem státního či obchodního tajemství

Forma vydání

tištěná verze "print"

Odkazy

URL

Označené pro přenos do RIV

Ano

Organizační jednotka

Fakulta informatiky

ISBN

979-8-4007-1504-4

DOI

https://doi.org/10.1145/3716863

Klíčová slova anglicky

Binary decision diagrams; Decision trees; Learning; Explainability; Decision making and control; Strategy synthesis

Štítky

firank_B

Změněno: 29. 1. 2026 15:57, Jana Halámková

Anotace

V originále

Safety-critical controllers of complex systems are hard to construct manually. Automated approaches such as controller synthesis or learning provide a tempting alternative but usually lack explainability. To this end, learning decision trees (DTs) has been prevalently used towards an interpretable model of the generated controllers. However, DTs do not exploit shared decision making, a key concept exploited in binary decision diagrams (BDDs) to reduce their size and thus improve explainability. In this work, we introduce predicate decision diagrams (PDDs) that extend BDDs with predicates and thus unite the advantages of DTs and BDDs for controller representation. We establish a synthesis pipeline for efficient construction of PDDs from DTs representing controllers, exploiting reduction techniques for BDDs also for PDDs.

Návaznosti

MUNI/I/1757/2021, interní kód MU

Název: MUNI Award in Science and Humanities (Akronym: Křetínský)

Investor: Masarykova univerzita, MUNI Award in Science and Humanities, MASH - MUNI Award in Science and Humanities

Citovat

CHAKRABORTY, Debraj; Clemens DUBSLAFF; Sudeep KANAV; Jan KŘETÍNSKÝ a Christoph WEINHUBER. Explaining Control Policies through Predicate Decision Diagrams. In HSCC '25: Proceedings of the 28th ACM International Conference on Hybrid Systems: Computation and Control. Irvine, USA: Association for Computing Machinery, 2025, s. 1-12. ISBN 979-8-4007-1504-4. Dostupné z: https://doi.org/10.1145/3716863.

@inproceedings{2551178,
   author = {Chakraborty, Debraj and Dubslaff, Clemens and Kanav, Sudeep and Křetínský, Jan and Weinhuber, Christoph},
   address = {Irvine, USA},
   booktitle = {HSCC '25: Proceedings of the 28th ACM International Conference on Hybrid Systems: Computation and Control},
   doi = {https://doi.org/10.1145/3716863},
   keywords = {Binary decision diagrams; Decision trees; Learning; Explainability; Decision making and control; Strategy synthesis},
   howpublished = {tištěná verze "print"},
   language = {eng},
   location = {Irvine, USA},
   isbn = {979-8-4007-1504-4},
   pages = {1-12},
   publisher = {Association for Computing Machinery},
   title = {Explaining Control Policies through Predicate Decision Diagrams},
   url = {https://dl.acm.org/doi/10.1145/3716863.3718049},
   year = {2025}
}

TY  - CONF
ID  - 2551178
AU  - Chakraborty, Debraj - Dubslaff, Clemens - Kanav, Sudeep - Křetínský, Jan - Weinhuber, Christoph
PY  - 2025
TI  - Explaining Control Policies through Predicate Decision Diagrams
PB  - Association for Computing Machinery
CY  - Irvine, USA
SN  - 9798400715044
KW  - Binary decision diagrams
KW  - Decision trees
KW  - Learning
KW  - Explainability
KW  - Decision making and control
KW  - Strategy synthesis
UR  - https://dl.acm.org/doi/10.1145/3716863.3718049
N2  - Safety-critical controllers of complex systems are hard to construct manually. Automated approaches such as controller synthesis or learning provide a tempting alternative but usually lack explainability. To this end, learning decision trees (DTs) has been prevalently used towards an interpretable model of the generated controllers. However, DTs do not exploit shared decision making, a key concept exploited in binary decision diagrams (BDDs) to reduce their size and thus improve explainability. In this work, we introduce predicate decision diagrams (PDDs) that extend BDDs with predicates and thus unite the advantages of DTs and BDDs for controller representation. We establish a synthesis pipeline for efficient construction of PDDs from DTs representing controllers, exploiting reduction techniques for BDDs also for PDDs.
ER  -

CHAKRABORTY, Debraj; Clemens DUBSLAFF; Sudeep KANAV; Jan KŘETÍNSKÝ a Christoph WEINHUBER. Explaining Control Policies through Predicate Decision Diagrams. In \textit{HSCC '25: Proceedings of the 28th ACM International Conference on Hybrid Systems: Computation and Control}. Irvine, USA: Association for Computing Machinery, 2025, s.~1-12. ISBN~979-8-4007-1504-4. Dostupné z: https://doi.org/10.1145/3716863.

Přehled o publikaci