Estimation of change point in failure rate models (Q1193945)

Estimation of the change point is considered for a failure rate function \(\gamma(x)\) which is an increasing decreasing failure rate or a decreasing increasing failure rate function (IDFR or DIFR). Kernel density estimators and empirical cumulative distribution functions are utilized for estimating \(f(x)\) and \(F(x)\), respectively, which are used to estimate \(\gamma(x)\) and its change point. Then asymptotic properties of the change point estimator are considered. Simulation results are given for different \(\gamma(x)\) and kernel functions.