cure-models

Survival models for insurance — cure models, CLV, lapse tables, MLflow wrapper

Mixture cure models — Weibull/LogNormal/Cox MCM, EM algorithm, Maller-Zhou identifiability, non-claimer scoring

Reproducible R implementation for assessing sufficient follow-up in cure models and applying extreme-value methods when follow-up is insufficient. Includes SFU hypothesis tests, EBVK tail-index estimation, real colon cancer data analysis, curse-of-dimensionality diagnostics, and comparisons with gradient-boosted GP regression.