By Dr. Rong Pan
Weibull distribution is widely used in failure time data analysis. In this three-hour webinar we will thoroughly discuss the Weibull regression model and its many applications in reliability modeling and data analysis problems. First, the properties of Weibull distribution, particularly the key properties that are related to failure time data, will be discussed. Second, we will demonstrate how to fit the failure data, including censored data, to a Weibull model and how to assess the goodness of fit. Next, the Weibull regression method, which is one of the most popular data analysis methods for reliability data with covariates such as the data from accelerated life tests, will be introduced. We will discuss how to evaluate the effect of a covariate and how to conduct residual analysis. An iterative model building process will be presented in an exercise. Finally, we will compare the Weibull regression model with the Cox’s proportional hazard model, another popular semi-parametric regression model.
In this webinar, we will focus on the statistical data analysis techniques for building failure time regression models and for predicting product reliability. In addition, based on the properties of reliability prediction, we can properly compare reliabilities of products from different manufacturers/suppliers and construct statistically efficient tests to demonstrate reliability. As nowadays most data analysis can be performed by computers, we will demonstrate the Weibull analysis by using Minitab® and JMP®, as well as some Excel® templates that are customized for specific reliability data analysis tasks.