LeapsAndBounds: A Method for Approximately Optimal Algorithm Configuration.
Gellért WeiszAndrás GyörgyCsaba SzepesváriPublished in: CoRR (2018)
Keyphrases
- high accuracy
- cost function
- k means
- dynamic programming
- recognition algorithm
- significant improvement
- optimization algorithm
- detection method
- improved algorithm
- experimental evaluation
- theoretical analysis
- objective function
- segmentation algorithm
- synthetic and real images
- computational cost
- computational complexity
- detection algorithm
- estimation algorithm
- preprocessing
- clustering method
- segmentation method
- high efficiency
- computationally efficient
- tree structure
- support vector machine svm
- classification algorithm
- matching algorithm
- classification method
- input data
- computational efficiency
- optimization method
- energy function
- np hard
- single pass
- convergence rate
- similarity measure
- selection algorithm
- reconstruction method
- combinatorial optimization
- expectation maximization
- simulated annealing
- worst case
- probabilistic model
- em algorithm
- optimal solution
- learning algorithm
- approximately optimal
- linear programming
- search space