2015/12/03 by Nikhil Padmanabhan, Martin White, Harrison H. Zhou +2
Computer Science · Mathematics · Physics and Astronomy · #Algorithm #Applied mathematics #Artificial intelligence #Computer science #Convergence (economics) #Galaxies: Formation, Evolution, Phenomena #Gaussian #Gaussian Processes and Bayesian Inference #Limit (mathematics) #Mathematical analysis #Mathematics #Matrix (chemical analysis) #Physics #Quantum mechanics #Rate of convergence #Regularization (linguistics) #Scale (ratio) #Sparse matrix #Statistical and numerical algorithms #Statistical physics #astro-ph.CO #astro-ph.IM #stat.ME
paper · pdf · doi:10.1093/mnras/stw1042
11 pages, 14 figures, submitted to MNRAS
arxiv created 2015/12/03 · openalex publication_date 2016/05/03 · arxiv updated 2016/05/25 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/05
We apply a method recently introduced to the statistical literature to directly estimate the precision matrix from an ensemble of samples drawn from a corresponding Gaussian distribution. Motivated by the observation that cosmological precision matrices are often approximately sparse, the method allows one to exploit this sparsity of the precision matrix to more quickly converge to an asymptotic |1/√N\rm sim| rate while simultaneously providing an error model for all of the terms. Such an estimate can be used as the starting point for further regularization efforts which can improve upon the |1/√N\rm sim| limit above, and incorporating such additional steps is straightforward within this framework. We demonstrate the technique with toy models and with an example motivated by large-scale structure two-point analysis, showing significant improvements in the rate of convergence. For the large-scale structure example, we find errors on the precision matrix which are factors of 5 smaller than for the sample precision matrix for thousands of simulations or, alternatively, convergence to the same error level with more than an order of magnitude fewer simulations.