Recent progress in technology has made large high-quality data sets widely accessible for practitioners. This opens up an interesting possibility of exploring Big models, but at the same time serious challenges are faced with traditional inference methodology. This is especially true for Bayesian inference, where simulation algorithms are often deemed to be too computationally expensive.
The idea of this session is to discuss our experience working with big data and developing scalable inferential algorithms. I can share my experience of Markov chain Monte Carlo in the presence of large datasets and some work we have done in this area.
Recent progress in technology has made large high-quality data sets widely accessible for practitioners. This opens up an interesting possibility of exploring Big models, but at the same time serious challenges are faced with traditional inference methodology. This is especially true for Bayesian inference, where simulation algorithms are often deemed to be too computationally expensive.
The idea of this session is to discuss our experience working with big data and developing scalable inferential algorithms. I can share my experience of Markov chain Monte Carlo in the presence of large datasets and some work we have done in this area.