Favoring the k-Means Algorithm with Initialization Methods.
Anderson Francisco de OliveiraMaria do Carmo NicolettiPublished in: ISDA (1) (2018)
Keyphrases
- k means
- preprocessing
- spectral clustering
- computational cost
- expectation maximization
- recently published
- significant improvement
- synthetic and real images
- detection algorithm
- clustering algorithm
- benchmark data sets
- data clustering
- clustering method
- dynamic programming
- optimization algorithm
- heuristic methods
- similarity measure
- self organizing maps
- parameter settings
- cost function
- synthetic and real datasets
- learning algorithm
- iterative algorithms
- noisy data
- computationally demanding
- test images
- matching algorithm
- fuzzy k means
- probabilistic model
- simulated annealing
- rough k means
- high dimensional
- cluster analysis
- experimental evaluation
- worst case