Machine learning based prediction of regularization parameters for seismic inverse problems

• Autorzy: Liu, Shihuan , Zhang, Jiashu
• Języki publikacji: EN

Abstrakty

EN

Regularization parameter selection (RPS) is one of the most important tasks in solving inverse problems. The most common approaches seek the optimal regularization parameter (ORP) from a sequence of candidate values. However, these methods are often time-consuming because they need to conduct the estimation process on all candidate values, and they are always restricted to solve certain problem types. In this paper, we propose a novel machine learning-based prediction framework (MLBP) for the RPS problem. The MLBP frst generates a large number of synthetic data by varying the inputs with diferent noise conditions. Then, MLBP extracts some pre-defned features to represent the input data and computes the ORP of each synthetic example by using true models. The pairs of ORP and extracted features construct a training set, which is used to train a regression model to describe the relationship between the ORP and input data. Therefore, for newly practical inverse problems, MLBP can predict their ORPs directly with the pre-trained regression model, avoiding wasting computational resources on improper regularization parameters. The numerical results also show that MLBP requires signifcantly less computing time and provides more accurate solutions for diferent tasks than traditional methods. Especially, even though the MLBP trains the regression model on synthetic data, it can also achieve satisfying performance when directly applied to feld data.

Słowa kluczowe

PL

parametr regularyzacji; wybór; nauczanie maszynowe; odwrotne problemy; inwersja sejsmiczna

EN

regularization parameter; selection; machine learning; inverse problems; seismic inversion

• Czasopismo: Acta Geophysica
• Rocznik: 2021
• Tom: Vol. 69, no. 3
• Strony: 809--820
• Opis fizyczny: Bibliogr. 44 poz.

Twórcy

autor

Liu, Shihuan

• Sichuan Province Key Lab of Signal and Information Processing, Southwest Jiaotong University, Chengdu 610031, China,
• School of Mathematics and Physics, Jinggangshan University, Ji’an 343009, China

autor

Zhang, Jiashu

• Sichuan Province Key Lab of Signal and Information Processing, Southwest Jiaotong University, Chengdu 610031, China,

Bibliografia

• 1. Ali J, Khan R, Ahmad N, Maqsood I (2012) Random forests and decision trees. Int J Comput Sci Issues (IJCSI) 9(5):272
• 2. Belgiu M, Drăguţ L (2016) Random forest in remote sensing: a review of applications and future directions. ISPRS J Photogr Remote Sens 114:24–31
• 3. Bonesky T (2008) Morozov’s discrepancy principle and tikhonov-type functionals. Inverse Problem 25(1):015015
• 4. Buland A, Omre H (2003) Bayesian linearized avo inversion. Geophysics 68(1):185–198
• 5. Calvetti D, Morigi S, Reichel L, Sgallari F (2000) Tikhonov regularization and the l-curve for large discrete ill-posed problems. J Comput Appl Math 123(1–2):423–446
• 6. Cootes TF, Ionita MC, Lindner C, Sauer P (2012) Robust and accurate shape model fitting using random forest regression voting. In: European Conference on Computer Vision, Springer, pp 278–291
• 7. Du S, Zhang J, Hu G (2020) A robust data-driven AVO inversion with logarithm absolute error loss function. Acta Geophysica 68(2):445–458
• 8. Gholami A (2016) A fast automatic multichannel blind seismic inversion for high-resolution impedance recovery. Geophysics 81(5):V357–V364
• 9. Gholami A, Sacchi MD (2013) Fast 3d blind seismic deconvolution via constrained total variation and gcv. SIAM J Imag Sci 6(4):2350–2369
• 10. Golub GH, Heath M, Wahba G (1979) Generalized cross-validation as a method for choosing a good ridge parameter. Technometrics 21(2):215–223
• 11. Hamid H, Pidlisecky A (2015) Multitrace impedance inversion with lateral constraints. Geophysics 80(6):M101–M111
• 12. Hamid H, Pidlisecky A, Lines L (2018) Prestack structurally constrained impedance inversion. Geophysics 83(2):R89–R103
• 13. Hansen PC (1992) Analysis of discrete ill-posed problems by means of the l-curve. SIAM Rev 34(4):561–580
• 14. Huang G, Li J, Luo C, Chen X (2017) Regularization parameter adaptive acquisition based on improved gcv method and its application in pre-stack avo inversion. In: SEG Technical Program Expanded Abstracts 2017, Society of Exploration Geophysicists, pp 748–752
• 15. Jiang K, Wang Z, Yi P, Wang G, Lu T, Jiang J (2019) Edge-enhanced gan for remote sensing image superresolution. IEEE Trans Geosci Remote Sens 57(8):5799–5812
• 16. Johnston PR, Gulrajani RM (2000) Selecting the corner in the l-curve approach to tikhonov regularization. IEEE Trans Biomed Eng 47(9):1293–1296
• 17. Langer A (2017) Automated parameter selection for total variation minimization in image restoration. J Math Imag Vision 57(2):239–268
• 18. Li C, Zhang F (2017) Amplitude-versus-angle inversion based on the l 1-norm-based likelihood function and the total variation regularization constraint. Geophysics 82(3):R173–R182
• 19. Li S, Gui J, Gao J, Wang S, Li H (2019) Direct inversion for sensitive elastic parameters of deep reservoirs. Acta Geophysica 67(5):1329–1340
• 20. Liao H, Ng MK (2010) Blind deconvolution using generalized cross-validation approach to regularization parameter estimation. IEEE Trans Image Process 20(3):670–680
• 21. Liaw A, Wiener M et al (2002) Classification and regression by randomforest. R News 2(3):18–22
• 22. Mboup M, Join C, Fliess M (2009) Numerical differentiation with annihilators in noisy environment. Num Algorith 50(4):439–467
• 23. Mostafaei K, Ramazi H (2018) Compiling and verifying 3d models of 2d induced polarization and resistivity data by geostatistical methods. Acta Geophysica 66(5):959–971
• 24. Nguyen N, Milanfar P, Golub G (2001) Efficient generalized cross-validation with applications to parametric image restoration and resolution enhancement. IEEE Trans Image Process 10(9):1299–1308
• 25. Nikolova M (2004) A variational approach to remove outliers and impulse noise. J Math Imag Vision 20(1–2):99–120
• 26. Papyan V, Elad M (2016) Multi-scale patch-based image restoration. IEEE Trans Image Process 25(1):249–261
• 27. Pérez DO, Velis DR, Sacchi MD (2013) High-resolution prestack seismic inversion using a hybrid fista least-squares strategy. Geophysics 78(5):R185–R195
• 28. Pérez DO, Velis DR, Sacchi MD (2017) Three-term inversion of prestack seismic data using a weighted l2, 1 mixed norm. Geophys Prospect 65(6):1477–1495
• 29. Qu S, Zhou H, Liu R, Chen Y, Zu S, Yu S, Yuan J, Yang Y (2016) Deblending of simultaneous-source seismic data using fast iterative shrinkage-thresholding algorithm with firm-thresholding. Acta Geophysica 64(4):1064–1092
• 30. She B, Wang Y, Liang J, Liu Z, Song C, Hu G (2018) A data-driven amplitude variation with offset inversion method via learned dictionaries and sparse representation. Geophysics 83(6):R725–R748
• 31. She B, Wang Y, Liang J, Hu G (2019) Data-driven simultaneous seismic inversion of multiparameters via collaborative sparse representation. Geophys J Int 218(1):313–332
• 32. Shen H, Peng L, Yue L, Yuan Q, Zhang L (2015) Adaptive norm selection for regularized image restoration and super-resolution. IEEE Trans Cyber 46(6):1388–1399
• 33. Smith PF, Ganesh S, Liu P (2013) A comparison of random forest regression and multiple linear regression for prediction in neuroscience. J Neurosci Methods 220(1):85–91
• 34. Svetnik V, Liaw A, Tong C, Culberson JC, Sheridan RP, Feuston BP (2003) Random forest: a classification and regression tool for compound classification and qsar modeling. J Chem Inform Comput Sci 43(6):1947–1958
• 35. Tikhonov A (1963) Solution of incorrectly formulated problems and the regularization method. Soviet Meth Dokl 4:1035–1038
• 36. Tourbier S, Bresson X, Hagmann P, Thiran JP, Meuli R, Cuadra MB (2015) An efficient total variation algorithm for super-resolution in fetal brain mri with adaptive regularization. NeuroImage 118:584–597
• 37. Versteeg R (1994) The marmousi experience: velocity model determination on a synthetic complex data set. Lead Edge 13(9):927–936
• 38. Wang Y, Lin W, Cheng S, She B, Hu G, Liu W (2018) Sharp and laterally constrained multitrace impedance inversion based on blocky coordinate descent. Acta Geophysica 66(4):623–631
• 39. Wen YW, Chan RH (2011) Parameter selection for total-variation-based image restoration using discrepancy principle. IEEE Trans Image Process 21(4):1770–1781
• 40. Wu X (2017) Structure-, stratigraphy-and fault-guided regularization in geophysical inversion. Geophys J Int 210(1):184–195
• 41. Yang J, Wright J, Huang TS, Ma Y (2010) Image super-resolution via sparse representation. IEEE Trans Image Process 19(11):2861–2873
• 42. Yuan S, Wang S, Ma M, Ji Y, Deng L (2017) Sparse bayesian learning-based time-variant deconvolution. IEEE Trans Geosci Remote Sens 55(11):6182–6194
• 43. Yuan S, Wang S, Luo Y, Wei W, Wang G (2019) Impedance inversion by using the low-frequency full-waveform inversion result as an a priori model. Geophysics 84(2):R149–R164
• 44. Zhang J, Lv S, Liu Y, Hu G (2013) Avo inversion based on generalized extreme value distribution with adaptive parameter estimation. J Appl Geophys 98:11–20

Identyfikatory

DOI

10.1007/s11600-021-00569-7