ozerbakar/forecasting-the-level-and-dispersion-of-mortality-a-deep-learning-approach

This is an R research replication package for comparing five neural and four stochastic mortality-forecasting approaches across national populations, sexes, and rolling forecast origins. The documented workflow rebuilds data from Human Mortality Database counts, validates inputs, fits both model families, scores forecasts for life expectancy and life disparity, and produces tables and figures. The README supports these workflow relationships; source excerpts were not provided, so internal call details beyond those documented are omitted.