Developing Prognostic Systems of Cancer Patients by Ensemble Clustering

Algorithm 1

Ensemble algorithm of clustering of cancer data.

(1) Given 𝑁, 𝐾₁, and 𝐾₂, run the PAM clustering method 𝑁 times with each 𝐾 randomly chosen from [𝐾₁,𝐾₂].

(2) Construct the pairwise dissimilarity measure dis(𝐱_𝑖,𝐱_𝑗) by using the (2).

(3) Cluster the 𝑛 objects by applying a linkage method and the dissimilarity measure dis(𝐱_𝑖,𝐱_𝑗) from Step 2.