An Mahalanobis distances based text clustering algorithm

Though traditional fuzzy partitional text clustering algorithms are one of the most widely used methods, they are only fit to detect spherical structural clusters because they are based on Euclidean distances. When the data set has high dimensions, the accuracy and efficiency will decrease. Focus on solving this problem, a Fuzzy Mahalanobis distances based text clustering algorithm was proposed. Otherwise, finding eigenvalue and eigenvectors of a symmetric matrix or computing pseudoinvertion were used to avoid the singular values problem when finding Mahalanobis distances. The numerical experimental results show the validity of the proposed methods.