A method of two-stage clustering learning based on improved DBSCAN and density peak algorithm.
Mingyang LiXinhua BiLimin WangXuming HanPublished in: Comput. Commun. (2021)
Keyphrases
- clustering method
- k means
- density based clustering algorithm
- improved algorithm
- learning algorithm
- high accuracy
- data clustering
- detection method
- cost function
- similarity measure
- clustering algorithm
- dynamic programming
- preprocessing
- optimization algorithm
- hierarchical clustering
- detection algorithm
- experimental evaluation
- segmentation method
- theoretical analysis
- distance metric
- spectral clustering
- expectation maximization
- unsupervised learning
- convergence rate
- high efficiency
- computational cost
- cluster analysis
- optimization method
- classification algorithm
- density based clustering
- recognition algorithm
- objective function
- significant improvement
- tree structure
- hierarchical clustering algorithm
- segmentation algorithm
- computational complexity
- support vector machine svm
- unsupervised manner
- information bottleneck
- matching algorithm
- input data
- cluster centroids
- fuzzy c means
- support vector machine
- similarity matrix
- spatial clustering
- synthetic and real datasets
- image segmentation
- instance level constraints
- unsupervised clustering
- clustering quality
- similarity function
- optimal solution
- active learning
- document clustering
- data points