Big Data vs. complex physical models: a scalable inference algorithm
share
The data torrent unleashed by current and upcoming instruments requires scalable analysis methods. Machine Learning approaches scale well. However, separating the instrument measurement from the physical effects of interest, dealing with variable errors, and deriving parameter uncertainties is usually an afterthought. Classic forward-folding analyses with Markov Chain Monte Carlo or Nested Sampling enable parameter estimation and model comparison, even for complex and slow-to-evaluate physical models. However, these approaches require independent runs for each data set, implying an unfeasible number of model evaluations in the Big Data regime. Here we present a new algorithm, collaborative nested sampling, for deriving parameter probability distributions for each observation. Importantly, in our method the number of physical model evaluations scales sub-linearly with the number of data sets, and we make no assumptions about homogeneous errors, Gaussianity, the form of the model or heterogeneity/completeness of the observations. Collaborative nested sampling has immediate application in speeding up analyses of large surveys, integral-field-unit observations, and Monte Carlo simulations.
READ FULL TEXT