Mixtures of spatial factor analyzers for tensor-variate data

A mixture of spatial factor analyzers (MSFA) is introduced to address the challenges of clustering high-dimensional spatial data.

By leveraging the underlying coordinate system, the proposed framework incorporates a flexible, spline-based spatial decay covariance structure that prevents parameter inflation as dimensionality increases.

To model non-spatial dependence, matrix variate factor analyzers are employed for further dimensionality reduction.

Parameter estimation is conducted via a variant of the expectation-maximization algorithm combined with a generalized least squares estimator.

The proposed models are explored in the context of tensor-variate data analysis, where simulation studies and applications to Raman spectroscopy and hyperspectral texture databases demonstrate their capacity to accurately infer and differentiate distinct spatial patterns.