Bayesian-OverDBC: A Bayesian Density-Based Approach for Modeling Overlapping Clusters

Algorithm 2

Bayesian-OverDBC.

Baysian Overlapping Density Based Clustering Algorithm (OverDBC)

Input: Expression Matrix () , Data model

Output: Bayesian overlap clusters, (membership-matrix).

New cluster may be merged based on () probability.

//phase 1

(1) compute transaction matrix ()

//phase 2

(2) Find Core genes based on Density and Closeness Centrality

(3) Add gene to Core genes () based on density and Cc relations.

//phase 3

(4) For  All in (Set of Core Object) Repeat:

(5) If    Start Local Search to find nearest neighbors , Save cluster .

    Else For   to

     Based above probability select one of these paths:

      if    then Start local search to construct new cluster .

      Else  invoke func_bound_over( ) and return results

     End of For

End of If