Alexander Litvinenko is presenting his collaboration work on Likelihood Approximation With Hierarchical Matrices For Large Spatial Datasets at 2018 SIAM PP conference in Tokyo Japan

Alexander Litvinenko is presenting his collaboration work at the SIAM PP conference in Tokyo, Japan, March 7-10, 2018.

This work is done between the stochastic numerics group, Extreme Computing Research Center, and two statistical groups (led by Prof. M. Genton and Prof. Y. Sun) at KAUST.
The work is devoted to developing fast and scalable algorithms for typical computations in spatial statistics, where large covariance matrices are involved. Examples are an approximation of the likelihood in the MLE methods, kriging, trace, log-determinant, Cholesky decomposition, and matrix inverse. You can read more in
 
1. A. Litvinenko, Y. Sun, M. Genton, and D. Keyes, Likelihood Approximation With Hierarchical Matrices For Large Spatial Datasets, submitted to J. Computational Statistics and Data Analysis, major revisions,  arXiv, 1709.04419, 2017

2. A. Litvinenko, HLIBCov: Parallel Hierarchical Matrix Approximation of Large Covariance Matrices and Likelihoods with Applications in Parameter Identification, submitted to J. Communications in Comput. Physics, arXiv 1709.08625, 2017