Hello Friends! I am happy to release this video on Kaplan-Meier (KM) nonparametric method to estimate reliability. The KM method is used in situations when the data cannot be modelled using mathematical distributions such as Weibull, Lognormal etc. The KM method is popularly used in medical science to estimate survival probabilities of patients with diseases such as cancer, kidney diseases, etc.
Nice to have Hemant’s article showing Kaplan-Meier estimator along with my article complaining about its validity.
Notice that Hemant’s inputs are individual lifetimes along with a censoring indicator “Failed?” That avoids the possibility of cohort sizes being nonstationary or random processes, which I complained about in my article in this week’s Weekly Update.
I am working on a sequel that will illustrate some random cohort processes and the nonparametric reliability estimators that deal with exceptions to Kaplan-Meier conditions for validity and provide a test for validity of the Kaplan-Meier estimator from periodic cohorts.