This week's seminars
Statistics Seminars: Higher order Markov processes for the failure process of a repairable system
21 May 2018 15:30 in CM221
Most commonly used models for the failure process of a repairable system have two drawbacks: (1) they assume that the system is composed of one component, and (2) they may contain too many unknown parameters that must be estimated from failure data. However, most real-world systems are multi-component systems and failure data are too sparse to obtain stable estimates for models with many parameters. This necessitates development of new models to overcome the drawbacks. This presentation introduces a higher order Markov process model and investigates its special case, both of which model the failure process of a repairable multi-component system and contain a small number of unknown parameters. We derive a parameter estimation method and compares the performance of the proposed models with nine other models based on artificially generated data and fifteen real-world datasets. The results show that the two new models outperform the nine models, respectively.
Contact email@example.com for more information