Hybrid DCOP Solvers: Boosting Performance of Local Search Algorithms
van Leeuwen, Cornelis Jan, Pawełczak, Przemyzław
–arXiv.org Artificial Intelligence
We propose a novel method for expediting both symmetric and asymmetric Distributed Constraint Optimization Problem (DCOP) solvers. The core idea is based on initializing DCOP solvers with greedy fast non-iterative DCOP solvers. This is contrary to existing methods where initialization is always achieved using a random value assignment. We empirically show that changing the starting conditions of existing DCOP solvers not only reduces the algorithm convergence time by up to 50\%, but also reduces the communication overhead and leads to a better solution quality. We show that this effect is due to structural improvements in the variable assignment, which is caused by the spreading pattern of DCOP algorithm activation.) /Subject (Hybrid DCOPs)
arXiv.org Artificial Intelligence
Sep-4-2020
- Country:
- South America > Brazil
- São Paulo (0.04)
- North America
- United States
- Texas > Travis County
- Austin (0.04)
- New York > New York County
- New York City (0.04)
- California > Los Angeles County
- Los Angeles (0.28)
- Arizona > Pima County
- Tucson (0.04)
- Texas > Travis County
- Canada > Ontario
- Toronto (0.04)
- United States
- Europe
- Portugal (0.04)
- United Kingdom > Scotland
- City of Edinburgh > Edinburgh (0.04)
- Spain > Valencian Community
- Valencia Province > Valencia (0.04)
- Netherlands > South Holland
- Delft (0.04)
- Italy > Emilia-Romagna
- Metropolitan City of Bologna > Bologna (0.04)
- France > Île-de-France
- Czechia > South Moravian Region
- Brno (0.04)
- South America > Brazil
- Genre:
- Research Report > Promising Solution (0.34)
- Technology: