Prof Frank Coolen: Nonparametric predictive inference in reliability
Introduction by Prof Tavner. During the last decade, I have been developing a novel statistical approach, called 'nonparametric predictive inference' (NPI), as an attempt to limit the mathematical assumptions underlying statistical inference, and in which uncertainty is quantified via lower and upper probabilities. NPI has been developed for several basic problems in reliability, and enables some attractive inferences beyond the possibilities provided by established statistical approaches. Examples are NPI for system reliability, including redundancy allocation, and NPI prediction of unobserved failure modes. In this talk, a brief introduction to NPI will be followed by some such examples.
Contact email@example.com for more information about this event.