Automatic K selection method for the K - Means algorithm.
Xiuli ShaoHuichao LeeYiwei LiuBo ShenPublished in: ICSAI (2017)
Keyphrases
- k means
- clustering method
- high accuracy
- detection algorithm
- dynamic programming
- improved algorithm
- selection algorithm
- computational cost
- cost function
- significant improvement
- preprocessing
- detection method
- synthetic and real images
- computational complexity
- objective function
- similarity measure
- optimization algorithm
- selection strategy
- segmentation algorithm
- optimization method
- high efficiency
- estimation algorithm
- computationally efficient
- theoretical analysis
- experimental evaluation
- segmentation method
- convergence rate
- support vector machine svm
- reconstruction method
- single pass
- initial cluster centers
- tree structure
- probabilistic model
- matching algorithm
- classification method
- fully automatic
- learning algorithm
- data clustering
- classification algorithm
- input data
- search space
- neural network
- hierarchical clustering
- region of interest
- clustering algorithm
- fuzzy clustering algorithm
- variable weighting
- decision trees
- feature selection
- fine tuning