Skip to main content Accessibility help
×
Home

A Compressed Sensing Approach for Partial Differential Equations with Random Input Data

  • L. Mathelin (a1) and K. A. Gallivan (a2)

Abstract

In this paper, a novel approach for quantifying the parametric uncertainty associated with a stochastic problem output is presented. As with Monte-Carlo and stochastic collocation methods, only point-wise evaluations of the stochastic output response surface are required allowing the use of legacy deterministic codes and precluding the need for any dedicated stochastic code to solve the uncertain problem of interest. The new approach differs from these standard methods in that it is based on ideas directly linked to the recently developed compressed sensing theory. The technique allows the retrieval of the modes that contribute most significantly to the approximation of the solution using a minimal amount of information. The generation of this information, via many solver calls, is almost always the bottle-neck of an uncertainty quantification procedure. If the stochastic model output has a reasonably compressible representation in the retained approximation basis, the proposed method makes the best use of the available information and retrieves the dominant modes. Uncertainty quantification of the solution of both a 2-D and 8-D stochastic Shallow Water problem is used to demonstrate the significant performance improvement of the new method, requiring up to several orders of magnitude fewer solver calls than the usual sparse grid-based Polynomial Chaos (Smolyak scheme) to achieve comparable approximation accuracy.

Copyright

Corresponding author

Corresponding author.Email:mathelin@limsi.fr

References

Hide All
[1]Abramowitz, M., and Stegun, I. Handbook of mathematical functions. Dover, 1970.
[2]Babusška, I., Nobile, F., and Tempone, R. A stochastic collocation method for elliptic partial differentialequations with random input data. SIAMJ. Numer. Anal. 45, 3 (2007), 10051034.
[3]Becker, S., Bobin, J., and Candès, , Nesta, E.: a fast and accurate first-order method for sparse recovery. Tech. rep., Caltech Institute of Technology, 2009.
[4]Blanchard, J., Cartis, C., and Tanner, J. Compressed sensing: how sharp is the restricted isometry property ? SIAM review (2010). to appear.
[5]Cai, T., Wang, L., and Xu, G.New bounds for restricted isometry constants. Tech. rep., Massachusetts Institute of Technology, 2009.
[6]Cai, T., Xu, G., and Zhang, J. On recovery of sparse signals via ℓ1 minimization. IEEE Trans. Inf. Theory 55 (2009), 33883397.
[7]Cameron, R., and Martin, W.The orthogonal development of non-linear fonctionals in series of fourier-hermite fonctionals. Ann. Math. 48, 2 (1947), 385392.
[8]Candès, E., and Plan, Y.Near-ideal model selection by ℓ1 minimization. Annals of Statistics 37 (2007), 21452177.
[9]Candès, E., and Romberg, J. Quantitative robust uncertainty principles and optimally sparse decompositions. Found. Comput. Math. 6, 2 (2006), 227254.
[10]Candès, E., Romberg, J., and Tao, T. Stable signal recovery from incomplete measurements. Comm. Pure Appl. Math. 59 (2005), 12071223.
[11]Candeès, E., and Tao, T.Near-optimal signal recovery from random projections: universal encoding strategies. IEEE Trans. Inform. Theory 52 (2004), 54065425.
[12]Candeès, E., Wakin, M. and Boyd, S.Enhancing sparsity by reweighted ℓ1 minimization. J. Fourier Anal. Appl. 17 (2007), 877905.
[13]Chen, S., Donoho, D., and Saunders, M.Atomic decomposition by basis pursuit. SIAM J. Sci. Comput. 20 (1999), 3361.
[14]Deb, M., Babusška, I., and Oden, J.T.. Solution of stochastic partial differential equations using galerkin finite element techniques. Comput. Methods Appl. Mech. Eng. 190, 48 (2001), 6359–6372.
[15]Donoho, D.Compressed sensing. IEEE Trans. Infor. Theo. 52, 4 (2006), 12891306.
[16]Donoho, D., Elad, M., and Temlyakov, V.Stable recovery of sparse overcomplete representations in the presence of noise. IEEE Trans. Infor. Theo. 52, 1 (2006), 618.
[17]Doostan, A. and Owhadi, H. A sparse approximation of partial differential equations with random inputs. presented at the SIAM Annual Meeting, Pittsburgh, PA, USA, July 2010.
[18]Douglas, C., Haase, G., and Iskandarani, M.An additive schwarz preconditioner for the spectral element ocean model formulation of the shallow water equations. Elec. Trans. Numer. Anal. 15 (2003), 1828.
[19]Efron, B. and Tibshirani, R.Improvements on cross-validation: the. 632+ bootstrap method. J. Amer. Stat. Assoc. 92, 438 (1997), 548560.
[20]Figueiredo, M.Nowak, R., and Wright, S. Gradient projection for sparse reconstruction: application to compressed sensing and other inverse problems. IEEE Journal of Selected Topics in Signal Processing 1 (2007), 586597.
[21]Foucart, S.A note on guaranteed sparse recovery via ℓ1-minimization. Applied and Computational Harmonic Analysis 29, 1 (2010), 97103.
[22]Frauenfelder, P., Schwab, C., and Todor, R. Finite elements for elliptic problems with stochastic coefficients. Comput. Meth. Appl. Mech. Engrg. 194, 2–5 (2005), 205228.
[23]Ganapathysubramanian, B., and Zabaras, N.Sparse grid collocation methods for stochastic natural convection problems. J. Comput. Phys. 225 (2007), 652685.
[24]Ghanem, R., and Spanos, P.Stochastic finite elements. A spectral approach, rev. ed. Springer Verlag, 1991. 222 p.
[25]Gilbert, J., and Lemaréchal, C. Some numerical experiments with variable-storage quasinewton algorithms. Math. Program. 45 (1989), 407435.
[26]Iskandarani, M., Haidvogel, D., and Boyd, J. A staggered spectral element model with application to the oceanic shallow water equations. Int. J. Num. Meth. Fluids 20, 5 (1995), 393414.
[27]Le Maître, O., Knio, O., Najm, H., and Ghanem, R.Uncertainty propagation using WienerHaar expansions. J. Comput. Phys. 197, 1 (2004), 2857.
[28]Le Maître, O., Najm, H., Ghanem, R., and Knio,O., Multi-resolution analysis of Wiener-type uncertainty propagation schemes. J. Comput. Phys. 197, 2 (2004), 502531.
[29]Mallat, S., and Zhang, Z.Matching pursuit in a time-frequency dictionary. IEEE Trans. Signal Proc. 41, 12 (1993), 33973415.
[30]Mathelin, L., and Gallivan, K.Uncertainty quantification for sparse solution of random pdes. presented at the SIAM Annual Meeting, Pittsburgh, PA, USA, July 2010.
[31]Mathelin, L., Hussaini, M., and Zang, T.Stochastic approaches to uncertainty quantification in CFD simulations. Num. Algo. 38, 1 (2005), 209239.
[32]Mathelin, L., and Le Maître, O.Dual-based a posteriori error estimate for stochastic finite element methods. Comm. App. Math. Comput. Sci. 2 (2007), 83115.
[33]Mathelin, L., Pastur, L., and Le Maître, O. A compressed-sensing approach for closed-loop optimal control of nonlinear systems. Theo. Comput. Fluid Dyn. (2011). In press, [DOI: 10.1007/s00162-011-0235-9].
[34]Nobile, F., Tempone, R., and Webster, C.An anisotropic sparse grid stochastic collocation method for partial differential equations with random input data. SIAM J. Numer. Anal. 46, 5 (2007), 24112442.
[35]Novak, E., and Ritter, K.Simple cubature formulas with high polynomial exactness. Constructive Approximation 15 (1999), 499522.
[36]Petras, K.Fast calculation in the smolyak algorithm. Num. Algo. 26 (2001), 93109.
[37]Rauhut, H.Compressive sensing and structured random matrices. In Theoretical Foundations and Numerical Methods for Sparse Recovery, M. Fornasier, Ed., Vol. 9 of Radon Series Comp. Appl. Math. deGruyter, 2010, pp. 192.
[38]Rauhut, H., and Ward, R. Sparse legendre expansions via ℓ1-minimization. preprint (2010).
[39]Smolyak, S.Quadrature and interpolation formulas for tensor products of certain classes of functions. Dokl. Akad. Nauk. SSSR 4 (1963), 240243.
[40]Sobol, I.Distribution of points in a cube and approximate evaluation of integrals. USSR Comput. Maths. Math. Phys. 7 (1967), 86112.
[41]Sobol, I.Uniformly distributed sequences with an additional uniform property. USSR Comput. Maths. Math. Phys. 16 (1977), 236242.
[42]Soize, C., and Ghanem, R.Physical systems with random uncertainties: chaos representations with arbitrary probability measure. SIAM J. Sci. Comput. 26, 2 (2004), 395410.
[43]Taylor, H., Bank, S., and McCoy, J.Deconvolution with the l 1-norm. Geophysics 44 (1979), 3952.
[44]Tibshirani, R.Regression shrinkage and selection via the lasso. J. Royal Statist. Soc. B 58 (1996), 267288.
[45]van den Berg, E., and Friedlander, MProbing the pareto frontier for basis pursuit solutions. SIAM J. Sci. Comput. 31, 2 (2008), 890912.
[46]Wan, X., and Karniadakis, G.An adaptive multi-element generalized polynomial chaos method for stochastic diffential equations. J. Comput. Phys. 209 (2005), 617642.
[47]Wiener, N.The homogeneous chaos. Amer. J. Math. 60, 4 (1938), 897936.
[48]Wright, S.Primal-Dual Interior-Point Methods. SIAM Publications, 1997.
[49]Xiu, D., and Hesthaven, J.High-order collocation methods for differential equations with random inputs. SIAM J. Sci. Comput. 27 (2005), 11181139.
[50]Xiu, D., and Karniadakis, G.The Wiener-Askey polynomial chaos for stochastic differential equations. SIAM J. Sci. Comput. 24, 2 (2002), 619644.
[51]Xiu, D., and Karniadakis, G.Modeling uncertainty in flow simulations via generalized polynomial chaos. J. Comput. Phys. 187 (2003), 137167.
[52]Zibulevsky, M., and Elad, M.L1-L2 optimization in signal and image processing. IEEE Signal Proc. Mag. 27, 3 (2010), 7688.

Keywords

Related content

Powered by UNSILO

A Compressed Sensing Approach for Partial Differential Equations with Random Input Data

  • L. Mathelin (a1) and K. A. Gallivan (a2)

Metrics

Full text views

Total number of HTML views: 0
Total number of PDF views: 0 *
Loading metrics...

Abstract views

Total abstract views: 0 *
Loading metrics...

* Views captured on Cambridge Core between <date>. This data will be updated every 24 hours.

Usage data cannot currently be displayed.