The k-means-u* algorithm: non-local jumps and greedy retries improve k-means++ clustering.
Bernd FritzkePublished in: CoRR (2017)
Keyphrases
- k means
- clustering algorithm
- clustering method
- probabilistic model
- learning algorithm
- expectation maximization
- computational cost
- preprocessing
- optimization algorithm
- theoretical analysis
- spectral clustering
- cluster analysis
- experimental evaluation
- computational complexity
- greedy heuristic
- particle swarm optimization
- computationally efficient
- dynamic programming
- improved algorithm
- greedy algorithm
- hierarchical clustering
- document clustering
- fuzzy k means
- unsupervised clustering
- clustering quality
- cluster centers
- globally optimal
- convergence rate
- similarity measure
- data clustering
- high accuracy
- optimal solution
- times faster
- matching algorithm
- association rules
- input data
- linear programming
- simulated annealing