Centroid 360: An Enhanced Centroid Initialization Method for K Means Algorithm.
Jovy Jay D. CabreraAriel M. SisonRuji P. MedinaPublished in: DSIT (2019)
Keyphrases
- k means
- clustering method
- clustering algorithm
- unsupervised clustering
- data clustering
- expectation maximization
- initial cluster centers
- high accuracy
- dynamic programming
- detection method
- experimental evaluation
- spectral clustering
- improved algorithm
- optimization algorithm
- cluster centers
- significant improvement
- recognition algorithm
- tree structure
- hierarchical clustering
- optimization method
- clustering quality
- synthetic and real images
- similarity measure
- center location
- learning algorithm
- computational complexity
- cluster analysis
- cost function
- preprocessing
- detection algorithm
- segmentation algorithm
- computationally efficient
- computational cost
- fuzzy k means
- input data
- single pass
- theoretical analysis
- convergence rate
- high efficiency
- selection algorithm
- clustering result
- hierarchical clustering algorithm
- fuzzy clustering algorithm
- matching algorithm
- document clustering