The probability density function of stochastic differential equations is governed by the Fokker-Planck (FP) equation. A novel machine learning method is developed to solve the general FP equations based on deep neural networks. The proposed algorithm does not require any interpolation and coordinate transformation, which is different from the traditional numerical methods. The main novelty of this paper is that penalty factors are introduced to overcome the local optimization for the deep learning approach, and the corresponding setting rules are given. Meanwhile, we consider a normalization condition as a supervision condition to effectively avoid that the trial solution is zero. Several numerical examples are presented to illustrate performances of the proposed algorithm, including one-, two-, and three-dimensional systems. All the results suggest that the deep learning is quite feasible and effective to calculate the FP equation. Furthermore, influences of the number of hidden layers, the penalty factors, and the optimization algorithm are discussed in detail. These results indicate that the performances of the machine learning technique can be improved through constructing the neural networks appropriately.
REFERENCES
1.
Z.
Wang
, Y.
Xu
, and H.
Yang
, “Lévy noise induced stochastic resonance in an FHN model
,” Sci. China Technol. Sci.
59
, 371
–375
(2016
). 2.
Y.
Li
, Y.
Xu
, J.
Kurths
, and X.
Yue
, “Lévy-noise-induced transport in a rough triple-well potential
,” Phys. Rev. E
94
, 042222
(2016
). 3.
M.
Escobedo
, M. A.
Herrero
, and J. J. L.
Velazquez
, “Radiation dynamics in homogeneous plasma
,” Physica D
126
, 236
–260
(1999
). 4.
A.
Ceccato
and D.
Frezzato
, “Remarks on the chemical Fokker-Planck and Langevin equations: Nonphysical currents at equilibrium
,” J. Chem. Phys.
148
, 064114
(2018
). 5.
B. M.
Grafov
, “Fokker-Planck equations for stochastic diffusion associated with Markovian electrochemical noise
,” Russ. J. Electrochem.
51
, 278
–280
(2015
). 6.
H.
Karcher
, S.
Lee
, M.
Kaazempurmofrad
, and R.
Kamm
, “A coarse-grained model for force-induced protein deformation and kinetics
,” Biophys. J.
90
, 2686
–2697
(2006
). 7.
Y.
Xu
, J.
Feng
, J.
Li
, and H.
Zhang
, “Lévy noise induced switch in the gene transcriptional regulatory system
,” Chaos
23
, 013110
(2013
). 8.
I. S. M.
Fokou
, C. N. D.
Buckjohn
, M. S.
Siewe
, and C.
Tchawoua
, “Probabilistic behavior analysis of a sandwiched buckled beam under Gaussian white noise with energy harvesting perspectives
,” Chaos Solitons Fractals
92
, 101
–114
(2016
). 9.
G.
Furioli
, A.
Pulvirenti
, E.
Terraneo
, and G.
Toscani
, “Fokker-Planck equations in the modeling of socio-economic phenomena
,” Math. Models Methods Appl. Sci.
27
, 115
–158
(2017
). 10.
J.
Elgin
, “The Fokker-Planck equation: Methods of solution and applications
,” Opt. Acta Int. J. Opt.
31
, 1206
–1207
(1996
). 11.
J.
Náprstek
and R.
Král
, “Finite element method analysis of Fokker-Planck equation in stationary and evolutionary versions
,” Adv. Eng. Softw.
72
, 28
–38
(2014
). 12.
R. F.
Galán
, E. G
Bard
, and N. N.
Urban
, “Stochastic dynamics of uncoupled neural oscillators: Fokker-Planck studies with the finite element method
,” Phys. Rev. E
76
, 056110
(2007
). 13.
Y.
Jiang
, “A new analysis of stability and convergence for finite difference schemes solving the time fractional Fokker-Planck equation
,” Appl. Math. Model.
39
, 1163
–1171
(2015
). 14.
B.
Sepehrian
and M. K.
Radpoor
, “Numerical solution of non-linear Fokker-Planck equation using finite differences method and the cubic spline functions
,” Appl. Math. Comput.
262
, 187
–190
(2015
). 15.
A. N.
Drozdov
, “Accurate path integral representations of the Fokker-Planck equation with a linear reference system: Comparative study of current theories
,” Phys. Rev. E
57
, 146
–158
(1998
). 16.
M.
Chandrika
and Y. N.
Kaznessis
, “Path-integral method for predicting relative binding affinities of protein-ligand complexes
,” J. Am. Chem. Soc.
131
, 4521
–4528
(2009
). 17.
Y.
Xu
, W.
Zan
, W.
Jia
, and J.
Kurths
, “Path integral solutions of the governing equation of SDEs excited by Lévy white noise
,” J. Comput. Phys.
394
, 41
–45
(2019
). 18.
J.
Biazar
, P.
Gholamin
, and K.
Hosseini
, “Variational iteration method for solving Fokker-Planck equation
,” J. Franklin Inst.
347
, 1137
–1147
(2010
). 19.
M.
Torvattanabun
and S.
Duangpithak
, “Numerical simulations of Fokker-Plank equation by variational iteration method
,” Int. J. Math. Anal.
5
, 2193
–2201
(2011
).20.
Z.
Sun
, J.
Carrillo
, and C. W.
Shu
, “A discontinuous Galerkin method for nonlinear parabolic equations and gradient flow problems with interaction potentials
,” J. Comput. Phys.
352
, 76
–104
(2017
). 21.
Y.
Zheng
, C.
Li
, and Z.
Zhao
, “A fully discrete discontinuous Galerkin method for nonlinear fractional Fokker-Planck equation
,” Math. Probl. Eng.
2010
, 279038
( 2010
). 22.
I.
Norio
and G.
Kosuke
, “Thermal fluctuations and stability of a particle levitated by a repulsive Casimir force in a liquid
,” Phys. Rev. E
88
, 052133
(2013
). 23.
E.
Hirvijoki
, T.
Kurki-Suonio
, S.
Äkäslompolo
, J.
Varje
, T.
Koskela
, and J.
Miettunen
, “Monte Carlo method and high performance computing for solving Fokker-Planck equation of minority plasma particles
,” J. Plasma Phys.
81
, 435810301
(2015
). 24.
D.
Barrow
and N.
Kourentzes
, “The impact of special days in call arrivals forecasting: A neural network approach to modelling special days
,” Eur. J. Oper. Res.
264
, 967
–977
(2018
). 25.
M.
Schirmer
, M.
Lehning
, and J.
Schweizer
, “Statistical forecasting of regional avalanche danger using simulated snow-cover data
,” J. Glaciol.
55
, 761
–768
(2017
). 26.
G.
Cybenko
, “Approximation by superpositions of a sigmoidal function
,” Math. Control Signals Syst.
2
, 303
–314
(1989
). 27.
I. E.
Lagaris
, A.
Likas
, and D. I.
Fotiadis
, “Artificial neural networks for solving ordinary and partial differential equations
,” IEEE Trans. Neural Netw.
9
, 987
–1000
(1998
). 28.
T.
Leephakpreeda
, “Novel determination of differential-equation solutions: Universal approximation method
,” J. Comput. Appl. Math.
146
, 443
–457
(2002
). 29.
B. P.
van Milligen
, V. V.
Tribaldos
, and J. A.
Jiménez
, “Neural network differential equation and plasma equilibrium solver
,” Phys. Rev. Lett.
75
, 3594
(1995
). 30.
R. S.
Beidokhti
and A.
Malek
, “Solving initial-boundary value problems for systems of partial differential equations using neural networks and optimization techniques
,” J. Franklin Inst.
346
, 898
–913
(2009
). 31.
A.
Malek
and R. S.
Beidokhti
, “Numerical solution for high order differential equations using a hybrid neural network-optimization method
,” Appl. Math. Comput.
183
, 260
–271
(2006
). 32.
J.
Berg
and K.
Nyström
, “A unified deep artificial neural network approach to partial differential equations in complex geometries
,” Neurocomputing
317
, 28
–41
(2018
). 33.
M.
Raissi
, P.
Perdikaris
, and G. E.
Karniadakis
, “Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
,” J. Comput. Phys.
378
, 686
–707
(2019
). 34.
J.
Han
, A.
Jentzen
, and E.
Weinan
, “Solving high-dimensional partial differential equations using deep learning
,” Proc. Natl. Acad. Sci. U.S.A.
115
, 8505
–8510
(2018
). 35.
J.
Sirignano
and K.
Spiliopoulos
, “DGM: A deep learning algorithm for solving partial differential equations
,” J. Comput. Phys.
375
, 1339
–1364
(2018
). 36.
C.
Zhu
, R. H.
Byrd
, P.
Lu
, and J.
Nocedal
, “Algorithm 778: L-BFGS-B: Fortran subroutines for large-scale bound-constrained optimization
,” ACM Trans. Math. Softw.
23
, 550
–560
(1997
). 37.
J.
Ma
, Y.
Xu
, Y.
Li
, R.
Tian
, and J.
Kurths
, “Predicting noise-induced critical transitions in bistable systems
,” Chaos
29
, 081102
(2019
). 38.
J.
Ma
, Y.
Xu
, W.
Xu
, Y.
Li
, and J.
Kurths
, “Slowing down critical transitions via Gaussian white noise and periodic force
,” Sci. China Technol. Sci.
62
, 2144
–2152
(2019
). 39.
R. L.
Honeycutt
, “Stochastic Runge-Kutta algorithms. I. White noise
,” Phys. Rev. A
45
, 600
(1992
). © 2020 Author(s).
2020
Author(s)
You do not currently have access to this content.
