Machine learning approximation algorithms for high-dimensional fully nonlinear partial differential equations and second-order backward stochastic differential equations

Abstract

High-dimensional partial differential equations (PDE) appear in a number of models from the financial industry, such as in derivative pricing models, credit valuation adjustment (CVA) models, or portfolio optimization models. The PDEs in such applications are high-dimensional as the dimension corresponds to the number of financial assets in a portfolio. Moreover, such PDEs are often fully nonlinear due to the need to incorporate certain nonlinear phenomena in the model such as default risks, transaction costs, volatility uncertainty (Knightian uncertainty), or trading constraints in the model. Such high-dimensional fully nonlinear PDEs are exceedingly difficult to solve as the computational effort for standard approximation methods grows exponentially with the dimension. In this work we propose a new method for solving high-dimensional fully nonlinear second-order PDEs. Our method can in particular be used to sample from high-dimensional nonlinear expectations. The method is based on (i) a connection between fully nonlinear second-order PDEs and second-order backward stochastic differential equations (2BSDEs), (ii) a merged formulation of the PDE and the 2BSDE problem, (iii) a temporal forward discretization of the 2BSDE and a spatial approximation via deep neural nets, and (iv) a stochastic gradient descent-type optimization procedure. Numerical results obtained using $\mathrm{T}\mathrm{ENSOR}\mathrm{F}\mathrm{LOW}$ in $\mathrm{P}\mathrm{YTHON}$ illustrate the efficiency and the accuracy of the method in the cases of a $100$-dimensional Black-Scholes-Barenblatt equation, a $100$-dimensional Hamilton-Jacobi-Bellman equation, and a nonlinear expectation of a $100$-dimensional $G$-Brownian motion.

56 pages, 12 figures

References in corpus (12)

• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• DGM: A deep learning algorithm for solving partial differential equations
• Deep learning-based numerical methods for high-dimensional parabolic partial differential equations and backward stochastic differential equations
• A regression-based Monte Carlo method to solve backward stochastic differential equations
• A proof that artificial neural networks overcome the curse of dimensionality in the numerical approximation of Black-Scholes partial differential equations
• On multilevel Picard numerical approximations for high-dimensional nonlinear parabolic partial differential equations and high-dimensional nonlinear backward stochastic differential equations
• Deep Learning Approximation for Stochastic Control Problems
• Strong error analysis for stochastic gradient descent optimization algorithms
• Stratified regression Monte-Carlo scheme for semilinear PDEs and BSDEs with large scale parallelization on GPUs
• Runge-Kutta schemes for backward stochastic differential equations
• Numerical simulation of BSDEs using empirical regression methods: theory and practice
• Variations on branching methods for non linear PDEs

Cited by in corpus (79)

• DeepXDE: A deep learning library for solving differential equations
• DGM: A deep learning algorithm for solving partial differential equations
• Surrogate Modeling for Fluid Flows Based on Physics-Constrained Deep Learning Without Simulation Data
• Physics-Constrained Deep Learning for High-dimensional Surrogate Modeling and Uncertainty Quantification without Labeled Data
• A Deep Collocation Method for the Bending Analysis of Kirchhoff Plate
• Weak Adversarial Networks for High-dimensional Partial Differential Equations
• Deep Autoencoder based Energy Method for the Bending, Vibration, and Buckling Analysis of Kirchhoff Plates
• A fast and accurate physics-informed neural network reduced order model with shallow masked autoencoder
• Tackling the Curse of Dimensionality with Physics-Informed Neural Networks
• Algorithms for Solving High Dimensional PDEs: From Nonlinear Monte Carlo to Machine Learning
• A proof that artificial neural networks overcome the curse of dimensionality in the numerical approximation of Black-Scholes partial differential equations
• Analysis of the Generalization Error: Empirical Risk Minimization over Deep Artificial Neural Networks Overcomes the Curse of Dimensionality in the Numerical Approximation of Black-Scholes Partial Differential Equations
• PFNN: A Penalty-Free Neural Network Method for Solving a Class of Second-Order Boundary-Value Problems on Complex Geometries
• Convergence of the Deep BSDE Method for Coupled FBSDEs
• Deep splitting method for parabolic PDEs
• Pricing options and computing implied volatilities using neural networks
• Stochastic analysis of heterogeneous porous material with modified neural architecture search (NAS) based physics-informed neural networks using transfer learning
• DNN Expression Rate Analysis of High-dimensional PDEs: Application to Option Pricing
• Multi-fidelity surrogate modeling using long short-term memory networks
• An overview on deep learning-based approximation methods for partial differential equations
• A proof that rectified deep neural networks overcome the curse of dimensionality in the numerical approximation of semilinear heat equations
• Solving the Kolmogorov PDE by means of deep learning
• Analysis of three dimensional potential problems in non-homogeneous media with physics-informed deep collocation method using material transfer learning and sensitivity analysis
• Solving forward and inverse problems of the logarithmic nonlinear Schrodinger equation with PT-symmetric harmonic potential via deep learning
• Numerical Solution of Inverse Problems by Weak Adversarial Networks
• On some neural network architectures that can represent viscosity solutions of certain high dimensional Hamilton--Jacobi partial differential equations
• Trainability and Accuracy of Neural Networks: An Interacting Particle System Approach
• Strong error analysis for stochastic gradient descent optimization algorithms
• Convergence Rate Analysis for Deep Ritz Method
• Overcoming the curse of dimensionality in the approximative pricing of financial derivatives with default risks
• Overcoming the curse of dimensionality in the numerical approximation of semilinear parabolic partial differential equations
• Overcoming the curse of dimensionality in the numerical approximation of Allen-Cahn partial differential equations via truncated full-history recursive multilevel Picard approximations
• Full error analysis for the training of deep neural networks
• Hutchinson Trace Estimation for High-Dimensional and High-Order Physics-Informed Neural Networks
• Numerical simulations for full history recursive multilevel Picard approximations for systems of high-dimensional partial differential equations
• Numerical methods for backward stochastic differential equations: A survey
• Efficient nonlinear manifold reduced order model
• Uniform error estimates for artificial neural network approximations for heat equations
• Mean-Field Langevin Dynamics and Energy Landscape of Neural Networks
• A probabilistic generative model for semi-supervised training of coarse-grained surrogates and enforcing physical constraints through virtual observables
• Exponential Convergence of Deep Operator Networks for Elliptic Partial Differential Equations
• Overcoming the curse of dimensionality in the numerical approximation of parabolic partial differential equations with gradient-dependent nonlinearities
• MIM: A deep mixed residual method for solving high-order partial differential equations
• Neural network architectures using min-plus algebra for solving certain high dimensional optimal control problems and Hamilton-Jacobi PDEs
• Deep neural network approximations for Monte Carlo algorithms
• Solving non-linear Kolmogorov equations in large dimensions by using deep learning: a numerical comparison of discretization schemes
• Neural networks-based algorithms for stochastic control and PDEs in finance
• Pricing equity-linked life insurance contracts with multiple risk factors by neural networks
• The Deep Ritz method: A deep learning-based numerical algorithm for solving variational problems
• Generalised multilevel Picard approximations
• Stability analysis of hierarchical tensor methods for time-dependent PDEs
• Enforcing exact boundary and initial conditions in the deep mixed residual method
• Deep learning based numerical approximation algorithms for stochastic partial differential equations
• Risk management with machine-learning-based algorithms
• Cell-average based neural network method for hyperbolic and parabolic partial differential equations
• A deep learning method for solving Fokker-Planck equations
• Monte Carlo for high-dimensional degenerated Semi Linear and Full Non Linear PDEs
• Overcoming the curse of dimensionality for some Hamilton--Jacobi partial differential equations via neural network architectures
• Neural Network Solutions to Differential Equations in Non-Convex Domains: Solving the Electric Field in the Slit-Well Microfluidic Device
• Learning to Discretize: Solving 1D Scalar Conservation Laws via Deep Reinforcement Learning
• Solving path dependent PDEs with LSTM networks and path signatures
• Quasi-Monte Carlo sampling for machine-learning partial differential equations
• Neural Network Approximation of Optimal Controls for Stochastic Reaction-Diffusion Equations
• Physics-informed neural networks for solving functional renormalization group on a lattice
• Deep learning for quadratic hedging in incomplete jump market
• Deep-learning based numerical BSDE method for barrier options
• A multiscale neural network based on hierarchical nested bases
• Novel multi-step predictor-corrector schemes for backward stochastic differential equations
• The Deep Parametric PDE Method: Application to Option Pricing
• Studying First Passage Problems using Neural Networks: A Case Study in the Slit-Well Microfluidic Device
• Numerical computation of probabilities for nonlinear SDEs in high dimension using Kolmogorov equation
• A learning scheme by sparse grids and Picard approximations for semilinear parabolic PDEs
• A Machine-Learning Method for Time-Dependent Wave Equations over Unbounded Domains
• Analytic Continuation of Noisy Data Using Adams Bashforth ResNet
• Design of A Two-point Steering Path Planner Using Geometric Control
• Semilinear Feynman-Kac Formulae for $B$-Continuous Viscosity Solutions
• Deep Learning for Constrained Utility Maximisation
• Quasi-Random Physics-informed Neural Networks
• Approximation of Optimal Feedback Controls for Stochastic Reaction-Diffusion Equations