A Survey of Dynamic Population Statistics and Applications in Test Evaluation

闫志强,蒋英杰,谢红卫
2009-01-01
Abstract:Dynamic population statistics is an important theory in test evaluation.After tracing its origin and developments,the primary problems,essential idea and general rules to solve them are discussed.Then,a more detailed survey is given to the applications and developments in evaluation of reliability growth test and weapon system performance.Finally,three leading methods to realize the dynamic population statistics are summarized and presented: multi-population synthetic estimation and inferences based on constraints relations,dynamic population modeling and forecasting based on linear models,and multi-source information fusion based on Bayesian method.
What problem does this paper attempt to address?