Clustering Method 2012



Tasks:

Study one of the following possible course project tasks and outline how the approach operates (or is going to be solved).
This choice does not need to be your course project work. You may also work as 2-person groups.

    - X-means algorithm (Pelleg&Moore, ICML, 2000)
    - Affinity propagation(Clustering by passing messages between data points, Frey&Dueck, Science, 2007), Find more information on this page.
    - DBSCAN algorithm (Sander,Ester,Kriegel&Xu, Data Mining & Knowledge Discovery, 1998)
    - Spectral Clustering , find demo here.
    - Cluster Ensemble (A.K. Jain, et al, A Combining multiple clusterings using evidence accumulation, TPAMI, 27(6), 2005 )
    - More efficient Genetic Algorithm (new) (A.K. Jain, et al, A Combining multiple clusterings using evidence accumulation, TPAMI, 27(6), 2005 )
    - K-means for Graph clustering (new) A reference
    - Random Swap for Graph clustering (new)
    - Greedy EM algorithm (GMM model), find matlab implementation here.
    - Another algorithm given later, or of your own (needs to be approved)
    - Create better and simpler web page for SIPU clustering data sets
    - Clustering animation (as in SIPU page) of Random Swap
    - Other task to be defined later.


Qinpei Zhao: email: zhao@cs.joensuu.fi / qinpei.zhao@uef.fi(B357)