Course project work tasks

1
Course project work tasks
Clustering Methods
Pasi Fränti

• Speech and Image Processing UnitSchool of
  Computing
• University of Eastern Finland

2
X-means

• X-means is a heuristic hierarchical method that
  tentatively splits every cluster and applies
  local k-means. Splits that provide improvement
  according to Bayesian information criterion are
  accepted.
• Kd-tree structure is used to speed-up k-means.
• The method is potentially faster and better than
  k-means but the improvements realize only with
  low dimensional data.

3
Affinity propagation
4
Spectral clustering
5
Cluster ensemble
6
Genetic algorithm next generation
7
K-means Random swap for graph clustering
8
Greedy EM algorithm
9
SIPU clustering data sets www-page
10
Clustering animation of Random Swap
11
SIPU clustering data sets www-page