042. Accelerated Failure Time Models in R

In this video we continue to analyze the customer churn data and demonstrate how the myriad of explanatory factors can be included in the model in order to better explain the behaviour. We show how to use standard regression techniques to calculate confidence intervals, test hypothesis, and do model selection, and point out some of the parametric diagnostic criteria which are available.
Video Timeline
00:00 - Introduction
01:52 - Baseline Model
13:54 - Residual Analysis
16:33 - Answering Questions of Interest

[@onyinkwaonyinkwa7083]

Good good work. quite informative

[@MeerkatStatistics]

Very nice lecture! At 15:50, it should be *thinner* tailed distribution.

[@isabelanieto5694]

Hello!!! I am very interested in understanding how to define the distribution of the error term in the AFT models. I understand there is a close connection between its distribution and the functional form of the hazard rate. Does the relationship go both ways? That is, if I know the funcional form of my hazard rate or of my survival function could I derive something for the distribution of my error term? SOS everywhere I read the distribution of u is assumed :/

[@ethanlv6980]

The course is great!
I want to try AFT model in R, where can find the code and data used in course?

[@etiniakpayang55]

Can an aft model use a linear mixed effect model as baseline distribution?