A New Algorithm for Generating Situation-Specific Bayesian Networks Using Bayes-Ball Method.
Laécio L. SantosRommel N. CarvalhoMarcelo LadeiraWeigang LiPublished in: URSW@ISWC (2016)
Keyphrases
- significant improvement
- high accuracy
- cost function
- bayesian networks
- detection algorithm
- computational cost
- dynamic programming
- theoretical analysis
- synthetic and real images
- detection method
- experimental evaluation
- optimization algorithm
- preprocessing
- improved algorithm
- computational efficiency
- optimization method
- input data
- computational complexity
- bayesian network structure
- objective function
- recognition algorithm
- k means
- high efficiency
- selection algorithm
- segmentation method
- classification method
- convergence rate
- support vector machine svm
- estimation algorithm
- generation method
- computationally efficient
- matching algorithm
- support vector machine
- clustering method
- reconstruction method
- similarity measure
- single pass
- tree structure
- classification algorithm
- pairwise
- neural network
- junction tree
- probability of correct classification
- prior information
- energy function
- segmentation algorithm
- em algorithm
- probabilistic model
- np hard
- optimal solution
- image segmentation
- expectation maximization
- probability distribution
- classification accuracy
- search space