Loading presentation...

Present Remotely

Send the link below via email or IM


Present to your audience

Start remote presentation

  • Invited audience members will follow you as you navigate and present
  • People invited to a presentation do not need a Prezi account
  • This link expires 10 minutes after you close the presentation
  • A maximum of 30 users can follow your presentation
  • Learn more about this feature in our knowledge base article

Do you really want to delete this prezi?

Neither you, nor the coeditors you shared it with will be able to recover it again.



No description

Chao Zhou

on 3 May 2010

Comments (0)

Please log in to add your comment.

Report abuse

Transcript of glm

Survival analysis Derivation Primary purpose: model and analyse time-to-
event data. Data that have a principal endpoint at the time when an event occurs. (failures)
Time until an electrical component fails, or time to first recurence of a tumor after initial treatment. Hazard function h(t) Censored observation
Exponential regression model Weibull regression model Cox proportional hazards model Compare with GLM Type 1: experiment stop at a particular time Type 2:experiment stop after r out of n failed Random censoring Left censored and doubly censored Interval censoring: event happend between two time points Truncation Nonparametric Methods Kaplan-Meier (K-M) estimate Empirical survivor function (esf): Treat censored data as if there were no censored observations. 12,13,13+,14,14,17+ 12,13,13,14,14,17 S(t) 6/6 5/6 3/6 1/6 0 t <12 12 13 14 17 t S(t)
<12 1
12 1*5/6 =5/6
13 5/6*3/5 =3/6
14 3/6*1/3 =1/6
17 1/6*0 =0 K-M adjusts the esf to reflect the presence of right-censored observations. t S(t)
<12 1
12 1*5/6=5/6
13 5/6*4/5=4/6
13+ 4/6*4/4=4/6
14 4/6*1/3=4/18
17+ 4/18*1/1=4/18 Tests related to nonparametric methods Compare two groups Fisher’s exact test: Mantel-Haenszel/log-rank test: Parametric Methods Weibull R code Work cited:
Survival Analysis Using S , by Mara Tableman and Jong Sung Kim;
Survival Analysis in R , by David Diez Thanks!
Full transcript