Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
reference for calibration by maximum likelihood for proportional hazards models.
0 views
Skip to first unread message
TFD
Nov 30, 2009, 4:24:16 PM11/30/09
to
I'm looking for a reference describing if and how good maximum
likelihood is for assessing model fit for a proportional hazards
model. I often choose the lowest AIC to pick the best model, but have
no firm understanding of how model fit to the data equates to fit of
observed vs predicted event times.
likelihood is for assessing model fit for a proportional hazards
model. I often choose the lowest AIC to pick the best model, but have
no firm understanding of how model fit to the data equates to fit of
observed vs predicted event times.
Thanks
0 new messages