I'm currently studying which models would best work in predicting age (continuous variable) from DNA-methylation microarrays. As, nowadays there are many extensions/variations becoming available based on (penalized) regression techniques, I was wondering which techniques are now out there that can be used. I've extensively looked for proper articles online, but I can't seem to find a good review of the latest techniques for applying regression to microarray data.
So if you know or if you are currently using a technique you favor, please post it here, so I can help create an overview of the contemporary methods. Thanks a lot.
P.S.: I know this topic resembles this post a lot Microarray Class Prediction - For Continuous Data? , but in this post only the obvious techniques are being mentioned and I try to do a deeper inquiry on the latest variations/extensions of the main regression techniques.
How about random survival forests?
Yeah nice idea. Seems to be an interesting candidate. And it has also got an R-package, so that's awesome too.
Actually, isn't a random survival forest only just suited for predicting time of survival of a population. Can it be transferred to age prediction, which seems to be something else?