A Software Defect Prediction Approach Based on Hybrid Feature Dimensionality Reduction

Algorithm 2

K-Medoids algorithm.

Input:

Original dataset, FC-Correlation vector: F, FF-Correlation matrix: S, The number of clusters: k, The maximum number of iterations: N

Output:

Clustering results of k clusters;

1. Select k features as the initial centroids of k clusters

2. Whilek centroids no longer change or reach the specified number of iterationsdo

3.  each feature is assigned to the centroids with the highest correlation.

4.  Update the centroids of k clusters

5. end while

6. return Clustering results of k clusters