ALGEBRE LINEAIRE AVANCEE
Objectives
Knowlegde of numerical methods that are efficient for the solution of large sparse linear systems of equations.
Understand the link between linear algebra and graph processing. Analyse the efficiency of a method with respect to complexity, computing time and memory footprint in the perspective of high performance computing.
Sparse linear algebra will be introduced and used to illustrate all these issues.
Know how to apply specific numerical methods to process matrices occuring in the area of data mining (i.e. non-negative factorization of matrices, partial linear least-square, graph partitioning, K-means clustering, multilinear algebra and tensors).
Bibliography
1/ J. Dongarra, I. Duff, D. Sorensen and H. van der Vorst, Solving Linear Systems on Vector and Shared Memory Computers, SIAM, 1991.
2/ I. Duff, A. Erisman and J.K. Reid. Direct Methods for Sparse Matrices, Second Edition, Oxford University Press, London, 2017.
3/ E. Estrada, M. Fox, G.-L. Oppo and D. J. Higham, Network Science: Complexity in Nature and Technology, Springer, 2010.