2000 character limit reached
An inversion problem for optical spectrum data via physics-guided machine learning
Published 3 Apr 2024 in physics.data-an, cond-mat.str-el, cs.LG, and physics.comp-ph | (2404.02387v1)
Abstract: We propose the regularized recurrent inference machine (rRIM), a novel machine-learning approach to solve the challenging problem of deriving the pairing glue function from measured optical spectra. The rRIM incorporates physical principles into both training and inference and affords noise robustness, flexibility with out-of-distribution data, and reduced data requirements. It effectively obtains reliable pairing glue functions from experimental optical spectra and yields promising solutions for similar inverse problems of the Fredholm integral equation of the first kind.
- Z. Phys. B 64, 189 (1986).
- Wu, M. K. et al. Superconductivity at 93 k in a new mixed-phase Y-Ba-Cu-O compound system at ambient pressure. Phys. Rev. Lett. 58, 908 (1987).
- Plakida, N. High-temperature cuprate superconductors (Springer-Verlag GmbH Berlin, Heidelberg, 2010).
- Allen, P. B. Electron-phonon effects in the infrared properties of metals. Phys. Rev. B 3, 305 (1971).
- Electronic states and optical spectra of htsc with electron-phonon coupling. Phys. C Supercond. its Appl. 178, 266 (1991).
- Inversion techniques for optical conductivity data. Phys. Rev. B 73, 184507 (2006).
- Evolution of the bosonic spectral density of the high-temperature superconductor Bi22{}_{2}start_FLOATSUBSCRIPT 2 end_FLOATSUBSCRIPTSr22{}_{2}start_FLOATSUBSCRIPT 2 end_FLOATSUBSCRIPTCaCu22{}_{2}start_FLOATSUBSCRIPT 2 end_FLOATSUBSCRIPTO8+δ8𝛿{}_{8+\delta}start_FLOATSUBSCRIPT 8 + italic_δ end_FLOATSUBSCRIPT. Phys. Rev. B 75, 144508 (2007).
- Wazwaz, A.-M. The regularization method for fredholm integral equations of the first kind. Computers and Mathematics with Applications 61, 2981–2986 (2011).
- Analytic continuation via domain knowledge free machine learning. Phys. Rev. B 98, 245101 (2018).
- Vapnik, V. N. Statistical Learning Theory (John Wiley and Sons, 1998).
- Dordevic, S. V. et al. Extracting the electron-boson spectral function α2F(ω)superscript𝛼2𝐹𝜔\alpha^{2}F(\omega)italic_α start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_F ( italic_ω ) from infrared and photoemission data using inverse theory. Phys. Rev. B 71, 104529 (2005).
- Hwang, J. et al. a𝑎aitalic_a-axis optical conductivity of detwinned ortho-ii YBa22{}_{2}start_FLOATSUBSCRIPT 2 end_FLOATSUBSCRIPTCu33{}_{3}start_FLOATSUBSCRIPT 3 end_FLOATSUBSCRIPTO6.506.50{}_{6.50}start_FLOATSUBSCRIPT 6.50 end_FLOATSUBSCRIPT. Phys. Rev. B 73, 014508 (2006).
- van Heumen, E. et al. Optical determination of the relation between the electron-boson coupling function and the critical temperature in high-Tcsubscript𝑇𝑐T_{c}italic_T start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT cuprates. Phys. Rev. B 79, 184512 (2009).
- Inverse problems: Tikhonov theory and algorithms, vol. 22 (World Scientific, 2014).
- Hwang, J. Intrinsic temperature-dependent evolutions in the electron-boson spectral density obtained from optical data. Sci. Rep. 6, 23647 (2016).
- Artificial neural network approach to the analytic continuation problem. Phys. Rev. Lett. 124, 056401 (2020).
- Tikhonov regularization and the l-curve for large discrete ill-posed problems. Journal of computational and applied mathematics 123, 423 (2000).
- Greedy tikhonov regularization for large linear ill-posed problems. International Journal of Computer Mathematics 84, 1151 (2007).
- A machine learning approach to optimal tikhonov regularization i: affine manifolds. Analysis and Applications 20, 353 (2022).
- Projected regression method for solving fredholm integral equations arising in the analytic continuation problem of quantum physics. Inverse Problems 33, 115007 (2017).
- Electron-boson spectral density functions of cuprates obtained from optical spectra via machine learning. Phys. Rev. B 104, 235154 (2021).
- Tnet: A model-constrained tikhonov network approach for inverse problems. arXiv preprint arXiv:2105.12033 (2021).
- Recurrent inference machines for solving inverse problems. arXiv preprint arXiv:1706.04008 (2017).
- Andrychowicz, M. et al. Learning to learn by gradient descent by gradient descent. In Proceedings of the 30th International Conference on Neural Information Processing Systems, NIPS’16, 3988 (Curran Associates Inc., Red Hook, NY, USA, 2016).
- Solving ill-posed inverse problems using iterative deep neural networks. Inverse Probl. 33, 124007 (2017).
- Morningstar, W. R. et al. Data-driven reconstruction of gravitationally lensed galaxies using recurrent inference machines. The Astrophysical Journal 883, 14 (2019).
- Landweber, L. An iteration formula for fredholm integral equations of the first kind. American Journal of Mathematics 73, 615 (1951).
- An overview of numerical methods for the first kind fredholm integral equation. SN Applied Sciences 1, 1178 (2019).
- Neumaier, A. Solving ill-conditioned and singular linear systems: A tutorial on regularization. SIAM Review 40, 636 (1998).
- Gradient projection for sparse reconstruction: Application to compressed sensing and other inverse problems. IEEE Journal of Selected Topics in Signal Processing 1, 586 (2007).
- Cho, K. et al. Learning phrase representations using RNN encoder-decoder for statistical machine translation. In Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP), 1724 (Association for Computational Linguistics, Doha, Qatar, 2014).
- Deep Learning (MIT Press, 2016).
- Adam: A method for stochastic optimization. In Bengio, Y. & LeCun, Y. (eds.) 3rd International Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015, Conference Track Proceedings (2015).
- Groetsch, C. W. Inverse problems in the mathematical sciences, vol. 52 (Springer, 1993).
- Compressed Sensing with Invertible Generative Models and Dependent Noise. CoRR abs/2003.08089 (2020).
- Ongie, G. et al. Deep learning techniques for inverse problems in imaging. IEEE J. Sel. Areas Inf. Theory 1, 39 (2020).
Sponsor
Paper Prompts
Sign up for free to create and run prompts on this paper using GPT-5.
Top Community Prompts
Collections
Sign up for free to add this paper to one or more collections.