学术报告(喻达磊 2024.1.16)
Model Averaging and Reliability Estimation of k-Out-of-n: G System with Model Uncertainty
摘要:Improving the reliability estimation of the k-out-of-n: G system provides support for its proper operation. Taking the example of a supply chain with the k-out-of-n: G system, we use the copula function to model the dependence structure among suppliers. A model averaging method is also introduced based on Kullback–Leibler (KL) loss to estimate the reliability of the supply chain with a k-out-of-n: G system. The proposed estimator accommodates the uncertainty in the dependence structures among suppliers and is the first to use the optimal model averaging for developing the reliability estimation. We prove the asymptotic optimality of the proposed estimator and the consistency of weights. Simulation studies and an example in analyzing a real dataset demonstrate the effectiveness of the proposed method.