Application of nuclear smoothing methods in the analysis of regional population mortality

Abstract

The article presents the results of a study of methods for one-dimensional estimation of probability distribution density based on core function and nuclear smoothing in the problem of mortality risk analysis for a cohort of patients with oncopathology. We used data from medical and sociological monitoring for 10 years of observation. As a result, the instantaneous risk function is estimated using the nuclear smoothing method using the Epanechnikov core function, which minimizes the average single quadratic error. The authors presented dependency graphs. The developed approach is applied to support management decisions in the municipal concept of active longevity.