040. Accelerated Failure Time Models
Автор: Dr. Dylan Spicker
Загружено: 2022-02-28
Просмотров: 6680
In this video we present the idea of accelerated failure time models which are the direct extension to the regression models discussed in the location-scale family. They are linear models on the log scale, and represent a re-scaling (multiplicatively) based on the baseline "normal" time to event. The error distribution defines a parametric model for the survival model, which can be used to fit even in the presence of censoring (as we had previously seen).
Video Timeline
00:00 - Introduction
01:06 - Accelerated Failure Time Models
04:05 - Properties of AFT Models
06:36 - Example: Log-Normal Distribution
Доступные форматы для скачивания:
Скачать видео mp4
-
Информация по загрузке: