Active Semisupervised Clustering Algorithm with Label Propagation for Imbalanced and Multidensity Datasets

Algorithm 3

𝐸𝑥𝑝𝑒𝑛𝑑(𝐶,𝑘_𝑎𝑣𝑔𝑑𝑖𝑠(𝐶),𝐷_𝑙).

(1) Get all the labeled data which belong to C from 𝐷_𝑙.

(2) Let 𝐷^𝑙_𝐶 denote these labeled data.

(3) Foreach  x in 𝐷^𝑙_𝐶  do

(4)  Compute density of data x, 𝑑𝑒𝑛𝑠𝑖𝑡𝑦(𝑥,𝐶)

(5) End Foreach

(6)  Sort 𝐷^𝑙_𝐶 in descending order according to data

     density

(7) While (𝐷^𝑙_𝐶 is not null)

(8)  Take out the first data x from 𝐷^𝑙_𝐶

(9)  Compute KNN(𝑥,𝐶)

(10)    Foreach  y in KNN(𝑥,𝐶)  do

(11)  If  𝑑𝑖𝑠_KNN𝑥,𝑦,𝐶≤𝑎𝑣𝑔𝑑𝑖𝑠

(12)    𝑦^𝑙←𝑥^𝑙

(13)   Insert y into 𝐷^𝑙_𝐶, and add y into 𝐷_𝑙

(14) End If

(15)    End Foreach

(16)    Delete x from 𝐷^𝑙_𝐶

(17) End while

(18) Return 𝐷_𝑙