Robust and Scalable SDE Learning: A Functional Perspective

Stochastic differential equations provide a rich class of flexible generative models, capable of describing a wide range of spatio-temporal processes. A host of recent work looks to learn data-representing SDEs, using neural networks and other flexible function approximators. Despite these advances, learning remains computationally expensive due to the sequential nature of SDE integrators. In this work, we propose an importance-sampling estimator for probabilities of observations of SDEs for the purposes of learning. Crucially, the approach we suggest does not rely on such integrators. The proposed method produces lower-variance gradient estimates compared to algorithms based on SDE integrators and has the added advantage of being embarrassingly parallelizable. This facilitates the effective use of large-scale parallel hardware for massive decreases in computation time.

[1]  Andreas Ruttor,et al.  Approximate Bayes learning of stochastic differential equations. , 2017, Physical review. E.

[2]  Anton Thalmaier,et al.  Stochastic Calculus of Variations in Mathematical Finance , 2005 .

[3]  Austin R. Benson,et al.  Neural Jump Stochastic Differential Equations , 2019, NeurIPS.

[4]  Arnulf Jentzen,et al.  On a perturbation theory and on strong convergence rates for stochastic ordinary and partial differential equations with nonglobally monotone coefficients , 2020 .

[5]  Patrick Kidger,et al.  Neural SDEs as Infinite-Dimensional GANs , 2021, ICML.

[6]  I. V. Girsanov On Transforming a Certain Class of Stochastic Processes by Absolutely Continuous Substitution of Measures , 1960 .

[7]  David Duvenaud,et al.  Neural Ordinary Differential Equations , 2018, NeurIPS.

[8]  Jonghun Park,et al.  Laplace based approximate posterior inference for differential equation models , 2016, Statistics and Computing.

[9]  K. McAuley,et al.  Bayesian Estimation in Stochastic Differential Equation Models via Laplace Approximation , 2016 .

[10]  Jason Yosinski,et al.  Hamiltonian Neural Networks , 2019, NeurIPS.

[11]  Maxim Raginsky,et al.  Neural Stochastic Differential Equations: Deep Latent Gaussian Models in the Diffusion Limit , 2019, ArXiv.

[12]  Kyle Cranmer,et al.  Hamiltonian Graph Networks with ODE Integrators , 2019, ArXiv.

[13]  Patrick Kidger,et al.  Neural Rough Differential Equations for Long Time Series , 2021, ICML.

[14]  Prashant Kumar,et al.  Stochastic turbulence modeling in RANS simulations via Multilevel Monte Carlo , 2018, Computers & Fluids.

[15]  Atsushi Yamashita,et al.  Graph Neural Ordinary Differential Equations , 2019, ArXiv.