FBSDE based neural network algorithms for high-dimensional quasilinear parabolic PDEs. (English) Zbl 07599601
Summary: In this paper, we propose forward and backward stochastic differential equations (FBSDEs) based deep neural network (DNN) learning algorithms for the solution of high dimensional quasi-linear parabolic partial differential equations (PDEs), which is related to the FBSDEs from the Pardoux-Peng theory. The algorithms rely on a learning process by minimizing the path-wise difference between two discrete stochastic processes, which are defined by the time discretization of the FBSDEs and the DNN representation of the PDE solution, respectively. The proposed algorithms are shown to generate DNN solution for a 100-dimensional Black-Scholes-Barenblatt equation, which is accurate in a finite region in the solution space, and has a convergence rate close to that of the Euler-Maruyama scheme used for discretizing the FBSDEs.
MSC:
| 60Hxx | Stochastic analysis |
| 65Nxx | Numerical methods for partial differential equations, boundary value problems |
| 35Qxx | Partial differential equations of mathematical physics and other areas of application |
Keywords:
forward and backward SDEs; Feynman-Kac formula; multi-scale deep neural network; quasilinear parabolic equations; Pardoux-Peng theorySoftware:
MscaleDNNReferences:
| [1] | Doob, J. L., Classical Potential Theory and Its Probabilistic Counterpart: Advanced Problems (Dec 2012), Springer Science & Business Media |
| [2] | Han, J.; Weinan, E., Deep learning approximation for stochastic control problems, (Deep Reinforcement Learning Workshop. Deep Reinforcement Learning Workshop, NIPS (2016)) |
| [3] | Han, J.; Jentzen, A.; Weinan, E., Solving high-dimensional partial differential equations using deep learning, Proc. Natl. Acad. Sci. USA, 115, 34, 8505-8510 (Aug 2018) · Zbl 1416.35137 |
| [4] | Han, J.; Long, J., Convergence of the deep BSDE method for coupled FBSDEs, Probab. Uncertain. Quant. Risk, 5, 1, 1-33 (Dec 2020) · Zbl 1454.60105 |
| [5] | Ji, S.; Peng, S.; Peng, Y.; Zhang, X., Three algorithms for solving high-dimensional fully coupled FBSDEs through deep learning, IEEE Intell. Syst., 35, 3, 71-84 (Feb 2020) |
| [6] | El Karoui, N.; Peng, S.; Quenez, M. C., Backward stochastic differential equations in finance, Math. Finance, 7, 1, 1-71 (Jan 1997) · Zbl 0884.90035 |
| [7] | Kloeden, P. E.; Platen, E., Numerical Solution of Stochastic Differential Equations (Apr 2013), Springer Science & Business Media |
| [8] | Karatzas, I.; Shreve, S. E., Brownian Motion and Stochastic Calculus, 47-127 (1998), Springer: Springer New York, NY |
| [9] | François-Lavet, V.; Henderson, P.; Islam, R.; Bellemare, M. G.; Pineau, J., An introduction to deep reinforcement learning (Nov 2018), arXiv preprint · Zbl 1448.68021 |
| [10] | Liu, Z. Q.; Cai, W.; Xu, Zhi-Qin John, Multi-scale deep neural network (MscaleDNN) for solving Poisson-Boltzmann equation in complex domains, Commun. Comput. Phys., 28, 5, 1970-2001 (2020) · Zbl 1473.65348 |
| [11] | Oksendal, B., Stochastic differential equations, (Stochastic Differential Equations (2003), Springer: Springer Berlin, Heidelberg), 65-84 · Zbl 1025.60026 |
| [12] | Pardoux, E.; Peng, S., Backward stochastic differential equations and quasilinear parabolic partial differential equations, (Stochastic Partial Differential Equations and Their Applications (1992), Springer: Springer Berlin, Heidelberg), 200-217 · Zbl 0766.60079 |
| [13] | Pavliotis, G.; Stuart, A., Multiscale Methods: Averaging and Homogenization (Jan 2008), Springer Science & Business Media · Zbl 1160.35006 |
| [14] | Raissi, M., Forward-backward stochastic neural networks: deep learning of high-dimensional partial differential equations (Apr 2018), arXiv preprint |
| [15] | Sutton, R. S.; Barto, A. G., Reinforcement Learning: An Introduction (Nov 2018), MIT Press · Zbl 1407.68009 |
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.
