Feature selection for k-means clustering stability: theoretical analysis and an algorithm.
Dimitrios MavroeidisElena MarchioriPublished in: Data Min. Knowl. Discov. (2014)
Keyphrases
- theoretical analysis
- learning algorithm
- feature selection
- times faster
- dynamic programming
- improved algorithm
- computational complexity
- k means
- cost function
- experimental evaluation
- high accuracy
- conditional mutual information
- optimization algorithm
- computationally efficient
- np hard
- preprocessing
- similarity measure
- worst case
- computational cost
- simulated annealing
- particle swarm optimization
- search space
- classification algorithm
- support vector
- spectral clustering
- feature extraction