Hello,
I have failure data on some devices. Of the 15000, 45 have failed, and the rest are censored at cycle 3000. Using the "Life Distribution" platform, I fit a model and can estimate DPPM at various processes.
I would like to randomly select 1/5 of the failed ones and repeat the analysis and do this, say 100 times, and then get a histogram of the estimated dppm at a particular cycle.
Is there an add-in SPSS or JMP pro tool to do this? This add-in will be useful for other analyses.
Thanks,
That is called k-fold sampling. In your case k = 5. If either of the two programs supports k-fold cross validation, that should do the trick.
You will find many examples of cross validation in R, for example: https://schmettow.github.io/New_Stats/wwm.html#cross-validation
- Badges
- Science topic
- Similar topics
- Philosophy
- Philosophy Of Science
- Reasoning