Logic rules meet deep learning: a novel approach for Ship Type classification

Manolis Pitsikalis; Thanh-Toan Do; Alexei Lisitsa; Shan Luo

doi:10.1007/978-3-030-91167-6_14

Logic rules meet deep learning: a novel approach for Ship Type classification

Manolis Pitsikalis, Thanh-Toan Do, Alexei Lisitsa, Shan Luo

Department of Data Science & AI

Research output: Chapter in Book/Report/Conference proceeding › Conference Paper › Research

6 Citations (Scopus)

Abstract

The shipping industry is an important component of the global trade and economy, however in order to ensure law compliance and safety it needs to be monitored. In this paper, we present a novel Ship Type classification model that combines vessel transmitted data from the Automatic Identification System, with vessel imagery. The main components of our approach are the Faster R-CNN Deep Neural Network and a Neuro-Fuzzy system with IF-THEN rules. We evaluate our model using real world data and showcase the advantages of this combination while also compare it with other methods. Results show that our model can increase prediction scores by up to 15.4% when compared with the next best model we considered, while also maintaining a level of explainability as opposed to common black box approaches.

Original language	English
Title of host publication	5th International Joint Conference, RuleML+RR 2021 Leuven, Belgium, September 13–15, 2021 Proceedings
Editors	Sotiris Moschoyiannis, Rafael Peñaloza, Jan Vanthienen, Ahmet Soylu, Dumitru Roman
Place of Publication	Cham Switzerland
Publisher	Springer
Pages	203-217
Number of pages	15
ISBN (Electronic)	9783030911676
ISBN (Print)	9783030911669
DOIs	https://doi.org/10.1007/978-3-030-91167-6_14
Publication status	Published - 2021
Event	International Joint Conference on Rules and Reasoning 2021 - Leuven, Belgium Duration: 13 Sept 2021 → 15 Sept 2021 Conference number: 5th https://link.springer.com/book/10.1007/978-3-030-91167-6 (Proceedings)

Publication series

Name	Lecture Notes in Computer Science
Publisher	Springer
Volume	12851
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	International Joint Conference on Rules and Reasoning 2021
Abbreviated title	RuleML+RR 2021
Country/Territory	Belgium
City	Leuven
Period	13/09/21 → 15/09/21
Internet address	https://link.springer.com/book/10.1007/978-3-030-91167-6 (Proceedings)

Keywords

Classification rules
Fuzzy rules
Object detection

Access to Document

10.1007/978-3-030-91167-6_14

Cite this

Pitsikalis, M., Do, T.-T., Lisitsa, A., & Luo, S. (2021). Logic rules meet deep learning: a novel approach for Ship Type classification. In S. Moschoyiannis, R. Peñaloza, J. Vanthienen, A. Soylu, & D. Roman (Eds.), 5th International Joint Conference, RuleML+RR 2021 Leuven, Belgium, September 13–15, 2021 Proceedings (pp. 203-217). (Lecture Notes in Computer Science; Vol. 12851). Springer. https://doi.org/10.1007/978-3-030-91167-6_14

Pitsikalis, Manolis ; Do, Thanh-Toan ; Lisitsa, Alexei et al. / Logic rules meet deep learning : a novel approach for Ship Type classification. 5th International Joint Conference, RuleML+RR 2021 Leuven, Belgium, September 13–15, 2021 Proceedings. editor / Sotiris Moschoyiannis ; Rafael Peñaloza ; Jan Vanthienen ; Ahmet Soylu ; Dumitru Roman. Cham Switzerland : Springer, 2021. pp. 203-217 (Lecture Notes in Computer Science).

@inproceedings{77dad7ead6274acaa16760f7c3807620,

title = "Logic rules meet deep learning: a novel approach for Ship Type classification",

abstract = "The shipping industry is an important component of the global trade and economy, however in order to ensure law compliance and safety it needs to be monitored. In this paper, we present a novel Ship Type classification model that combines vessel transmitted data from the Automatic Identification System, with vessel imagery. The main components of our approach are the Faster R-CNN Deep Neural Network and a Neuro-Fuzzy system with IF-THEN rules. We evaluate our model using real world data and showcase the advantages of this combination while also compare it with other methods. Results show that our model can increase prediction scores by up to 15.4\% when compared with the next best model we considered, while also maintaining a level of explainability as opposed to common black box approaches.",

keywords = "Classification rules, Fuzzy rules, Object detection",

author = "Manolis Pitsikalis and Thanh-Toan Do and Alexei Lisitsa and Shan Luo",

note = "Funding Information: Acknowledgements. This work has been funded by the Engineering and Physical Sciences Research Council (EPSRC) Centre for Doctoral Training in Distributed Algorithms at the University of Liverpool, and Denbridge Marine Limited (https://www. denbridgemarine.com), United Kingdom. Publisher Copyright: {\textcopyright} 2021, Springer Nature Switzerland AG.; International Joint Conference on Rules and Reasoning 2021, RuleML+RR 2021 ; Conference date: 13-09-2021 Through 15-09-2021",

year = "2021",

doi = "10.1007/978-3-030-91167-6\_14",

language = "English",

isbn = "9783030911669",

series = "Lecture Notes in Computer Science",

publisher = "Springer",

pages = "203--217",

editor = "Sotiris Moschoyiannis and Rafael Pe{\~n}aloza and Jan Vanthienen and Ahmet Soylu and Dumitru Roman",

booktitle = "5th International Joint Conference, RuleML+RR 2021 Leuven, Belgium, September 13–15, 2021 Proceedings",

address = "Switzerland",

url = "https://link.springer.com/book/10.1007/978-3-030-91167-6",

}

Pitsikalis, M, Do, T-T, Lisitsa, A & Luo, S 2021, Logic rules meet deep learning: a novel approach for Ship Type classification. in S Moschoyiannis, R Peñaloza, J Vanthienen, A Soylu & D Roman (eds), 5th International Joint Conference, RuleML+RR 2021 Leuven, Belgium, September 13–15, 2021 Proceedings. Lecture Notes in Computer Science, vol. 12851, Springer, Cham Switzerland, pp. 203-217, International Joint Conference on Rules and Reasoning 2021, Leuven, Belgium, 13/09/21. https://doi.org/10.1007/978-3-030-91167-6_14

Logic rules meet deep learning: a novel approach for Ship Type classification. / Pitsikalis, Manolis; Do, Thanh-Toan; Lisitsa, Alexei et al.
5th International Joint Conference, RuleML+RR 2021 Leuven, Belgium, September 13–15, 2021 Proceedings. ed. / Sotiris Moschoyiannis; Rafael Peñaloza; Jan Vanthienen; Ahmet Soylu; Dumitru Roman. Cham Switzerland: Springer, 2021. p. 203-217 (Lecture Notes in Computer Science; Vol. 12851).

Research output: Chapter in Book/Report/Conference proceeding › Conference Paper › Research

TY - GEN

T1 - Logic rules meet deep learning

T2 - International Joint Conference on Rules and Reasoning 2021

AU - Pitsikalis, Manolis

AU - Do, Thanh-Toan

AU - Lisitsa, Alexei

AU - Luo, Shan

N1 - Conference code: 5th

PY - 2021

Y1 - 2021

N2 - The shipping industry is an important component of the global trade and economy, however in order to ensure law compliance and safety it needs to be monitored. In this paper, we present a novel Ship Type classification model that combines vessel transmitted data from the Automatic Identification System, with vessel imagery. The main components of our approach are the Faster R-CNN Deep Neural Network and a Neuro-Fuzzy system with IF-THEN rules. We evaluate our model using real world data and showcase the advantages of this combination while also compare it with other methods. Results show that our model can increase prediction scores by up to 15.4% when compared with the next best model we considered, while also maintaining a level of explainability as opposed to common black box approaches.

AB - The shipping industry is an important component of the global trade and economy, however in order to ensure law compliance and safety it needs to be monitored. In this paper, we present a novel Ship Type classification model that combines vessel transmitted data from the Automatic Identification System, with vessel imagery. The main components of our approach are the Faster R-CNN Deep Neural Network and a Neuro-Fuzzy system with IF-THEN rules. We evaluate our model using real world data and showcase the advantages of this combination while also compare it with other methods. Results show that our model can increase prediction scores by up to 15.4% when compared with the next best model we considered, while also maintaining a level of explainability as opposed to common black box approaches.

KW - Classification rules

KW - Fuzzy rules

KW - Object detection

UR - https://www.scopus.com/pages/publications/85121868894

U2 - 10.1007/978-3-030-91167-6_14

DO - 10.1007/978-3-030-91167-6_14

M3 - Conference Paper

AN - SCOPUS:85121868894

SN - 9783030911669

T3 - Lecture Notes in Computer Science

SP - 203

EP - 217

BT - 5th International Joint Conference, RuleML+RR 2021 Leuven, Belgium, September 13–15, 2021 Proceedings

A2 - Moschoyiannis, Sotiris

A2 - Peñaloza, Rafael

A2 - Vanthienen, Jan

A2 - Soylu, Ahmet

A2 - Roman, Dumitru

PB - Springer

CY - Cham Switzerland

Y2 - 13 September 2021 through 15 September 2021

ER -

Pitsikalis M, Do TT, Lisitsa A, Luo S. Logic rules meet deep learning: a novel approach for Ship Type classification. In Moschoyiannis S, Peñaloza R, Vanthienen J, Soylu A, Roman D, editors, 5th International Joint Conference, RuleML+RR 2021 Leuven, Belgium, September 13–15, 2021 Proceedings. Cham Switzerland: Springer. 2021. p. 203-217. (Lecture Notes in Computer Science). doi: 10.1007/978-3-030-91167-6_14

Logic rules meet deep learning: a novel approach for Ship Type classification

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Cite this