In this paper we present the results of an empirical study of stochastic projection and stochastic gradient descent methods as means of obtaining approximate inverses and preconditioners for iterative methods. Results of numerical experiments are used to analyse scalability and overall suitability of the selected methods as practical tools for treatment of large linear systems of equations. The results are preliminary due to the code being not yet fully optimized.
CITATION STYLE
Şahin, M. E., Lebedev, A., & Alexandrov, V. (2020). Empirical analysis of stochastic methods of linear algebra. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 12143 LNCS, pp. 539–549). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-030-50436-6_40
Mendeley helps you to discover research relevant for your work.