Final Report Summary - BIGBAYES (Rich, Structured and Efficient Learning of Big Bayesian Models)
With complex, heterogeneous and large scale data becoming ever more common, we need sophisticated statistical and machine learning models that allow us to learn as much as possible and as efficiently as possible from such data. This project explored the use of nonparametric models for such purposes, and with diverse challenges ranging from theoretical understanding, to modelling and algorithmic developments, to software systems allowing ease of use by end users. The BigBayes project tackles these diverse challenges head on, and it has been broadly successful in advancing the state of the art across these challenges.
One of the most exciting developments that surfaced throughout the project was the merging of ideas from Bayesian nonparametrics and deep learning
One of the most exciting developments that surfaced throughout the project was the merging of ideas from Bayesian nonparametrics and deep learning