A Distributed Algorithm for the Cluster-Based Outlier Detection Using Unsupervised Extreme Learning Machines

Algorithm 1

UELM algorithm [8].

Input. The training data: .

Output. The label vector of cluster corresponding to

() (a) Construct the graph Laplacian of ;

() (b) Generate a pair of random values for each hidden neuron,

  and calculate the output matrix ;

() (c)

() if then

()   Find the generalized eigenvectors of Equation (11). Let

    .

() else

()   Find the generalized eigenvectors of Equation (13). Let

    ;

() (d) Calculate the embedding matrix: ;

() (e) Treat each row of as a point, and cluster the points into

   clusters using the -means algorithm. Let be the label

   vector of cluster index for all the points.

() return ;