RobMixReg

RobMixReg implements robust mixture regression methods for identifying and modeling latent subgroups and unobserved heterogeneity in high-dimensional predictor settings while accommodating outliers and diverse regression forms.

Key Features:

• Robustness: Handles outliers to stabilize parameter estimates and inference in the presence of anomalous data points.
• Flexibility: Supports diverse regression forms to allow model specification tailored to different data-generating relationships.
• High Dimensionality: Manages high-dimensional predictor spaces for analyses with large numbers of covariates.
• Comprehensive Integration: Combines robustness, flexible regression forms, and high-dimensional modeling within a mixture regression framework.

Scientific Applications:

• Latent subgroup analysis: Identifies and characterizes latent subgroups and unobserved heterogeneity within populations.
• Genomics: Applies to genomics for detecting heterogeneity and subgroup structure in high-dimensional omics data.
• Epidemiology: Applies to epidemiology for stratifying populations and modeling heterogeneous responses in observational studies.
• Social sciences: Applies to social sciences for detecting latent classes and heterogeneous effects across subpopulations.

Methodology:

Implements finite mixture models to approximate general distribution functions in a semi-parametric manner and capture unobserved heterogeneity.