A direct adaptive method for faster backpropagation learning: the RPROP algorithm.
Martin A. RiedmillerHeinrich BraunPublished in: ICNN (1993)
Keyphrases
- preprocessing
- detection algorithm
- improved algorithm
- cost function
- clustering method
- optimization algorithm
- high accuracy
- detection method
- estimation algorithm
- objective function
- computational cost
- k means
- dynamic programming
- significant improvement
- computational complexity
- segmentation algorithm
- theoretical analysis
- computational efficiency
- synthetic and real images
- support vector machine svm
- mathematical model
- convergence rate
- input data
- experimental evaluation
- high efficiency
- highly efficient
- computationally efficient
- classification method
- single pass
- optimization method
- matching algorithm
- combinatorial optimization
- similarity measure
- experimental study
- decomposition method
- region of interest
- classification algorithm
- segmentation method
- filtering algorithm
- evolutionary algorithm
- optimal solution
- reconstruction method
- recursive least squares
- decision directed
- kalman filter
- clustering algorithm
- np hard
- probabilistic model
- simulated annealing
- neural network
- learning algorithm
- levenberg marquardt
- pairwise
- em algorithm
- selection algorithm
- recognition algorithm
- expectation maximization
- particle swarm optimization
- support vector machine
- tree structure
- noisy data
- prior information
- image matching
- test images