Pattern Matching & Extrapolation

[jwbradley87]

Hi everyone, 

 

I'm new here! I hope I've found the right place to post this. 

 

My question is this... Is there a specific mathematical field of study around the following problem? 

 

I have a collection of data points that describe a population. For any individual member of that population, how can I use the data points for that individual, in the context of the full data set, to extrapolate and predict the missing data points for the individual?

 

 

This fits into a larger concept that I've been toying with recently, so I'd really like to have a conversation with someone who understands the models/algorithms etc that could be used to solve the above problem. If you have some spare time on your hands too, and want to get involved in something that I think might be quite exciting, even better!

 

 

Looking forward to your replies

 

James

[hausinthehouse]

I'm only an entering graduate student, but I'm fairly sure this is a pretty big topic in biostats, at least.

[cyberwulf]

What you're describing is what we refer to in statistics as imputation. Work on imputation and related missing data methods ranges from the fairly theoretical (more often done in statistics departments) to the more applied (more often done in biostatistics departments). A good starting point would be a book by two 'giants' in the field, Little & Rubin's Statistical Analysis with Missing Data