Finding Hamiltonian Cycles with Graph Neural Networks

We train a small message-passing graph neural network to predict Hamiltonian cycles on Erdos-Renyl random graphs in a critical regime. It outperforms existing hand-crafted heuristics after about 2.5 hours of training on a single GPU. Our findings encourage an alternative approach to solving computat...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:2023 International Symposium on Image and Signal Processing and Analysis (ISPA) s. 1 - 6
Hlavní autoři: Bosnic, Filip, Sikic, Mile
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 18.09.2023
Témata:
Graph algorithms
Graphics processing units
Heuristics design
Machine learning
Machine learning algorithms
Neural nets
NP-hard problem
Signal processing
Signal processing algorithms
Supervised learning
Training
ISSN:1849-2266
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:We train a small message-passing graph neural network to predict Hamiltonian cycles on Erdos-Renyl random graphs in a critical regime. It outperforms existing hand-crafted heuristics after about 2.5 hours of training on a single GPU. Our findings encourage an alternative approach to solving computationally demanding (NP-hard) problems arising in practice. Instead of devising a heuristic by hand, one can train it end-to-end using a neural network. This has several advantages. Firstly, it is relatively quick and requires little problem-specific knowledge. Secondly, the network can adjust to the distribution of training samples, improving the performance on the most relevant problem instances. The model is trained using supervised learning on artificially created problem instances; this training procedure does not use an existing solver to produce the super-vised signal. Finally, the model generalizes well to larger graph sizes and retains reasonable performance even on graphs eight times the original size.
ISSN:1849-2266
DOI:10.1109/ISPA58351.2023.10278690