On estimation of hazard rate and reversed hazard rate based on transformations

Reliability concepts and measures are heavily used in the modern world to investigate characterists of lifetime data. Hazard rate, mean residual life, reversed hazard rate and expected inactivity time are four highly regarded tools that are used in the process. They have been studied in depth over the years. This paper proposes a simple approach that makes use of relationships that exist between them to estimate these functions. The method involves fixed-design local polynomial estamators of reliability functions and their derivatives. Numercial examples show that this simple method provide good estimates for two popular parametric models in reliability.