epidemiology

Apr
4
comment How to create a covariate × time interaction term in Cox regression?
It might be helpful for others if you posted a link to that resource here.
Mar
23
awarded Yearling
Mar
23
awarded Yearling
Mar
21
comment Confidence intervals for cross-validated statistics
Well, for each fold in each repetition of the cross-validation resampling, we not only calculated the maximum likelihood estimate of our statistic of interest, but also its confidence limits. We then took the average over all these replications for each the lower and upper confidence bound as well as the main estimate. In effect we cross-validated the confidence limits like any other statistic.
Mar
21
revised Confidence intervals for cross-validated statistics
deleted 1 characters in body
Mar
21
revised Can independent variables with low correlation with dependent variable be significant predictors?
edited body
Mar
21
accepted What can be inferred when multivariable ordinary least squares and quantile (median) regression yield differing results?
Mar
20
answered Can independent variables with low correlation with dependent variable be significant predictors?
Mar
12
comment Confidence intervals for cross-validated statistics
We went with cross-validating the confidence limits. It was computationally expensive.
Mar
10
asked What can be inferred when multivariable ordinary least squares and quantile (median) regression yield differing results?
Feb
24
asked Selecting features and estimating their out-of-sample performance with cross-validation
Jan
28
comment how does Cox proportional hazards model deal with time-dependent variables
Quite the opposite. Using only the baseline values can introduce bias if your predictor changes over time.
Jan
28
comment Is there a way to use cross validation to do variable/feature selection in R?
I think your procedure is unlikely to improve on the LASSO, whose implementations in R (e.g. glmnet and penalized) do by default employ cross-validation to find the "optimal" regularization parameter. One thing you could consider is repeating the LASSO search for this parameter several times to cope with the potentially large variance of cross-validation (repeated CV). Of course, no algorithm can beat your subject-specific prior knowledge.
Jan
14
comment how does Cox proportional hazards model deal with time-dependent variables
This is a broad question. The cox model evaluates the covariable values at all time points at which events occur. Longer intervals should have, all other things equal, a higher chance of containing more time points at which events occur, and therefore the covariable values associated with them a higher influence on the estimated hazard ratio.
Jan
6
awarded Popular Question
Dec
25
asked Tuning a model for predictive performance in a narrow probability range
Nov
15
comment What is the correct way to test for significant differences between coefficients?
If one wanted to test whether an effect is different between more than two groups, would an ANOVA comparing the model $y_i = \alpha + \beta x_i + \gamma g_i + \varepsilon_i$ and the one shown in this answer, $y_i = \alpha + \beta x_i + \gamma g_i + \delta (x_i \times g_i) + \varepsilon_i$ be appropriate?
Nov
4
accepted Fitting restricted cubic splines in a cox model
Oct
4
asked Sampling distribution when testing an estimated hazard ratio against a population HR when also estimating its SE
Sep
25
comment How can a combination of model parameters have a lower standard error than each individual coefficient?
Also the intuition is that if the covariance is negative, then if one is high, the other tends to be low, which makes less likely the occurence of extreme values in their sum, for which both would need to be either high or low. Right?
1 2 3 4 5