On the potential of normalized TSP features for automated algorithm selection.
Jonathan HeinsJakob BossekJanina PohlMoritz SeilerHeike TrautmannPascal KerschkePublished in: FOGA (2021)
Keyphrases
- np hard
- selection algorithm
- dynamic programming
- search space
- high accuracy
- cost function
- detection algorithm
- significant improvement
- optimization algorithm
- preprocessing
- times faster
- learning algorithm
- extracted features
- ant colony optimization
- genetic algorithm
- test images
- traveling salesman problem
- linear programming
- computational complexity
- optimal solution
- matching algorithm
- neural network
- matching process
- image segmentation
- feature subset
- convergence rate
- objective function
- combinatorial optimization
- search algorithm
- optimization problems
- clustering method
- evolutionary algorithm
- experimental evaluation
- segmentation algorithm
- theoretical analysis
- particle swarm optimization
- computational cost