ABSTRACT
This chapter focuses on the finite mixture models that have been used in high-dimensional settings and on the approaches that have been used for fitting and inference for these models. It describes motivating data examples are introduced as examples of high-dimensional continuous, categorical and mixed data. The chapter provides some issues involved in modeling high-dimensional data and discusses models for high-dimensional continuous data, categorical data and mixed data. It reviews some variable selection approaches for mixture modeling. The chapter explores the motivating examples are investigated using mixture models and shows that mixture models are able to appropriately model substructures in these data. Modeling the data using a mixture model will help establish if the clinical indicators suggest the existence of subgroups within the population, whether the subgroups correspond to the established pain types. One approach to modeling high-dimensional data using mixture models is to incorporate data reduction to reduce the data dimension.
