A novel feature-based approach to characterize algorithm performance for the traveling salesperson problem.
Olaf MersmannBernd BischlHeike TrautmannMarkus WagnerJakob BossekFrank NeumannPublished in: Ann. Math. Artif. Intell. (2013)
Keyphrases
- computational complexity
- high accuracy
- matching algorithm
- theoretical analysis
- learning algorithm
- times faster
- improved algorithm
- k means
- experimental evaluation
- detection algorithm
- simulated annealing
- worst case
- computational cost
- estimation algorithm
- single pass
- computationally efficient
- optimal solution
- dynamic programming
- objective function
- neural network
- recognition algorithm
- convergence rate
- convex hull
- preprocessing stage
- path planning
- tree structure
- optimization algorithm
- expectation maximization
- image features
- probabilistic model
- multi objective