Sequential selection for accelerated life testing via approximate Bayesian inference

Accelerated life testing (ALT) is typically used to assess the reliability of material's lifetime under desired stress levels. Recent advances in material engineering have made a variety alternatives readily available. To identify most reliable setting with efficient experimental design, sequential test planning strategy preferred. guarantee tractable statistical mechanism for information collection and update, we develop explicit model parameter update formulas via approximate Bayesian inference. Theories show that our give consistent estimates. Simulation study case proposed selection approach can significantly improve probability identifying alternative best performance over other design approaches.