TY - GEN
T1 - Machine-Learning Based Objective Function Selection for Community Detection
AU - Bornstein, Asa
AU - Rubin, Amir
AU - Hendler, Danny
N1 - Funding Information:
This work was supported in part by the Cyber Security Research Center at Ben-Gurion University.
Publisher Copyright:
© 2022, Springer Nature Switzerland AG.
PY - 2022/6/23
Y1 - 2022/6/23
N2 - NECTAR, a Node-centric ovErlapping Community deTection AlgoRithm, presented by Cohen et al., chooses dynamically between two objective functions which to optimize, based on the network on which it is invoked. It was shown that this approach outperforms six state-of-the-art algorithms for overlapping community detection. In this work, we present NECTAR-ML, an extension of the NECTAR algorithm that uses a machine-learning based model for automating the selection of the objective function, trained and evaluated on a dataset of 15,755 synthetic and 7 real-world networks. Our analysis shows that in approximately 90% of the cases our model was able to successfully select the correct objective function. We conducted a competitive analysis of NECTAR and NECTAR-ML. NECTAR-ML was shown to significantly outperform NECTAR’s ability to select the best objective function. We also conducted a competitive analysis of NECTAR-ML and two additional state-of-the-art multi-objective evolutionary community detection algorithms. NECTAR-ML outperformed both algorithms in terms of average detection quality. Multi-objective evolutionary algorithms are considered to be the most popular approach to solve multi-objective optimization problems and the fact that NECTAR-ML significantly outperforms them demonstrates the effectiveness of ML-based objective function selection.
AB - NECTAR, a Node-centric ovErlapping Community deTection AlgoRithm, presented by Cohen et al., chooses dynamically between two objective functions which to optimize, based on the network on which it is invoked. It was shown that this approach outperforms six state-of-the-art algorithms for overlapping community detection. In this work, we present NECTAR-ML, an extension of the NECTAR algorithm that uses a machine-learning based model for automating the selection of the objective function, trained and evaluated on a dataset of 15,755 synthetic and 7 real-world networks. Our analysis shows that in approximately 90% of the cases our model was able to successfully select the correct objective function. We conducted a competitive analysis of NECTAR and NECTAR-ML. NECTAR-ML was shown to significantly outperform NECTAR’s ability to select the best objective function. We also conducted a competitive analysis of NECTAR-ML and two additional state-of-the-art multi-objective evolutionary community detection algorithms. NECTAR-ML outperformed both algorithms in terms of average detection quality. Multi-objective evolutionary algorithms are considered to be the most popular approach to solve multi-objective optimization problems and the fact that NECTAR-ML significantly outperforms them demonstrates the effectiveness of ML-based objective function selection.
KW - Community detection
KW - Complex networks
KW - Machine learning
KW - Overlapping community detection
KW - Supervised learning
UR - http://www.scopus.com/inward/record.url?scp=85134153881&partnerID=8YFLogxK
U2 - 10.1007/978-3-031-07689-3_10
DO - 10.1007/978-3-031-07689-3_10
M3 - Conference contribution
AN - SCOPUS:85134153881
SN - 9783031076886
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 135
EP - 152
BT - Cyber Security, Cryptology, and Machine Learning - 6th International Symposium, CSCML 2022, Proceedings
A2 - Dolev, Shlomi
A2 - Meisels, Amnon
A2 - Katz, Jonathan
PB - Springer Science and Business Media Deutschland GmbH
T2 - 6th International Symposium on Cyber Security Cryptography and Machine Learning, CSCML 2022
Y2 - 30 June 2022 through 1 July 2022
ER -