| Abstract : | In survival analysis, the idea of hazard function is might be of more interest than the probability density function to a patient who had survived a certain time period and wanted to know something about its survival progress. Thus, in clinical trials the attention is focused on the estimation of hazard function and not in p.d.f.It is easily understood that a hazard function can not be always increasing or decreasing, which happen in many known survival models. It is necessary to befind parametric models which can describe the behavior of the hazard function properly. Most of the distributions that are used very common for the analysis ofsurvival data (e.g. Weibull and exponential) are not able to give hazard functions with exible shapes.
|
|---|
Copyright © 2013 Library AUEB. All rights reserved.