A new initialisation method for k-means algorithm in the clustering problem: data analysis.
Abolfazl KazemiGhazaleh KhodabandehlouiePublished in: Int. J. Data Anal. Tech. Strateg. (2018)
Keyphrases
- k means
- clustering method
- cluster analysis
- clustering algorithm
- data clustering
- spectral clustering
- cost function
- initial cluster centers
- cluster centers
- hierarchical clustering
- high accuracy
- computational complexity
- similarity measure
- unsupervised clustering
- detection algorithm
- computational cost
- preprocessing
- significant improvement
- clustering analysis
- improved algorithm
- dynamic programming
- clustering result
- data analysis
- agglomerative hierarchical clustering
- parameter free
- optimization method
- objective function
- clustering quality
- similarity matrix
- clustering approaches
- document clustering
- hierarchical clustering algorithm
- fuzzy clustering algorithm
- segmentation method
- variable weighting
- cluster structure
- classical clustering algorithms
- genetic k means algorithm
- input data
- similarity function
- learning algorithm
- matching algorithm
- tree structure
- fuzzy c means
- rough k means
- affinity propagation
- self organizing maps
- detection method
- probabilistic model
- expectation maximization
- optimization algorithm
- cluster centroids
- clustering ensemble
- cluster ensemble
- synthetic and real datasets
- fuzzy clustering