Novel data condensing method using a prototype's front propagation algorithm.
J. A. Pérez-BenitezJ. L. Pérez-BenítezJ. H. Espina-HernandezPublished in: Eng. Appl. Artif. Intell. (2015)
Keyphrases
- input data
- noisy data
- preprocessing
- high accuracy
- dynamic programming
- synthetic data
- experimental evaluation
- improved algorithm
- detection method
- computational complexity
- k means
- objective function
- clustering method
- computational cost
- optimization algorithm
- recognition algorithm
- detection algorithm
- data sets
- similarity measure
- synthetic and real images
- cost function
- significant improvement
- distance matrix
- computationally efficient
- tree structure
- information loss
- prior information
- energy function
- optimization method
- dissimilarity matrix
- convergence rate
- classification method
- segmentation method
- prior knowledge
- segmentation algorithm
- theoretical analysis
- missing data
- labeling process
- large scale data sets
- classification algorithm
- similarity matrix
- dimensional data
- extracted features
- estimation algorithm
- optimal solution
- sample set
- classification trees
- spectral clustering
- support vector machine svm
- data points
- probabilistic model
- feature space
- cluster centers
- missing values
- data structure
- training samples
- training data
- support vector machine
- expectation maximization
- particle swarm optimization