Robust and Flexible Model-Based Clustering for Data Sets with Missing Values at Random
Summary for MixtureMissing
MixtureMissing Plotting
Binary Classification Evaluation
Print for MixtureMissing
Mixture Model Selection
Extractor function for MixtureMissing
Missing-Data Pattern Generation
Missing Values Generation
Cluster Initialization using a Heuristic Method
Multivariate Contaminated Normal Mixture (MCNM)
Mean Imputation
Multivariate Generalized Hyperbolic Mixture (MGHM)
Implementations of various robust and flexible model-based clustering methods for data sets with missing values at random. Two main models are: Multivariate Contaminated Normal Mixture (MCNM, Tong and Tortora, 2022, <doi:10.1007/s11634-021-00476-1>) and Multivariate Generalized Hyperbolic Mixture (MGHM, Wei et al., 2019, <doi:10.1016/j.csda.2018.08.016>). Mixtures via some special or limiting cases of the multivariate generalized hyperbolic distribution are also included: Normal-Inverse Gaussian, Symmetric Normal-Inverse Gaussian, Skew-Cauchy, Cauchy, Skew-t, Student's t, Normal, Symmetric Generalized Hyperbolic, Hyperbolic Univariate Marginals, Hyperbolic, and Symmetric Hyperbolic.