Big Data Analysis

The last few decades have witness an explosion in model, theory and algorithm for big data analysis. Our lab is in the frontier of this research areas by developing foundational model and theory for big data analysis. Our general directions include developing dividing and conquering strategies for big data analysis, exploiting lower-dimensional strategy for high-dimensional data, and estimating and hypothesis testing of large covariance matrices. 


Related Publications 

Lee, K. and Lin, L. (2018).  Bayesian test and selection for bandwidth of high-dimensional Banded precision matrices. Submitted. Link.

Lee, K., Lee. J. and Lin, L. (2018). Minimax posterior convergence rates and model selection consistency in high-dimensional DAG models based on sparse Cholesky factors. Revision submitted.

Lee, K., Chae, M. and Lin, L. (2018). Bayesian high-dimensional semi-parametric inference beyond sub-Gaussian errors. Revision.

Chae, M., Lin, L and Dusnon, D.B. (2018) Bayesian sparse linear models with unknown symmetric errors. Information and Inference, revision submitted. Link.

Thomas, B., Lin, L., Lim, L., and Mukherjee, S (2018). Learning subspaces of different dimensions. Journal of Machine Learning Research. Submitted. Link.

Zhang, M., Lam, H. and Lin, L. (2018).  Robust and scalable Bayesian model selection. Computational Statistics & Data Analysis, Vol. 127, 229--247. Link.

Minsker, S., Srivastava, S., Lin, L. and Dunson, D.B. (2017). Robust and scalable Bayes via a median of subset posterior measure.  Journal of Machine Learning Research, 18(124):1--40. Link.