Skip to Main Content
The Possibilistic Latent Variable (PLV) clustering algorithm is a powerful tool for the analysis of complex datasets due to its robustness toward data distributions of different types and its ability to accurately identify the inherent clusters within the data. The scaling coefficient in the PLV algorithm plays a key role in reducing the effects of noise, thereby improving the precision of the clustering results. However, the optimal value of the scaling parameter varies depending on the population type of dataset. Accordingly, the current study proposes an evaluation method for evaluating suitable values of the scaling parameter. The relative comparison of each method is then examined by conducting PLV clustering trials using datasets comprising data of different types and patterns.