Efficient Nonparametric Population Modeling for Large Data Sets

In the context of biomedical data analysis, population models are used to characterize the average and individual behavior of a population of subjects. When a mechanistic model is not available, one can resort to the nonparametric approach that describes the individual curves as realizations of Gaussian processes. In this paper, efficient algorithms are developed for estimating the average and individual curves from large data sets collected in standardized experiments. The overall identification scheme presents a "client-server" architecture. The server takes care of managing historical information on past experiments. The client deals with a single new experiment and interrogates the server to obtain the information needed to reconstruct the individual curve. In this way, clients exploit the global data set without having access to the historical data and with negligible computational effort.