Algorithms for Model-Based Gaussian Hierarchical Clustering
SIAM Journal on Scientific Computing
Hi-index | 0.00 |
This paper presents a model-based Gaussian clustering approach for the identification of the porosity, hydrated and unhydrated phases constituting a backscatter electron image of the microstructure of cement paste. The likelihood function for the Gaussian mixture is optimised using the expectation-maximization algorithm and the quality of fit is assessed using the Bayes information criterion. The technique provides a consistent and repeatable means of phase identification within the microstructure of cement paste. The application of the approach has been demonstrated through an example.