Corpus ID: 165163523

Neural Stochastic Differential Equations: Deep Latent Gaussian Models in the Diffusion Limit

@article{Tzen2019NeuralSD,
  title={Neural Stochastic Differential Equations: Deep Latent Gaussian Models in the Diffusion Limit},
  author={Belinda Tzen and M. Raginsky},
  journal={ArXiv},
  year={2019},
  volume={abs/1905.09883}
}
  • Belinda Tzen, M. Raginsky
  • Published 2019
  • Computer Science, Mathematics
  • ArXiv
  • In deep latent Gaussian models, the latent variable is generated by a time-inhomogeneous Markov chain, where at each time step we pass the current state through a parametric nonlinear map, such as a feedforward neural net, and add a small independent Gaussian perturbation. This work considers the diffusion limit of such models, where the number of layers tends to infinity, while the step size and the noise variance tend to zero. The limiting latent object is an Ito diffusion process that solves… CONTINUE READING
    Deterministic Inference of Neural Stochastic Differential Equations
    Neural Jump Stochastic Differential Equations
    27
    Theoretical Guarantees for Learning Conditional Expectation using Controlled ODE-RNN
    1
    Stochasticity in Neural ODEs: An Empirical Study
    2
    Neural Optimal Control for Representation Learning
    Stochastic Normalizing Flows
    2
    Scalable Gradients for Stochastic Differential Equations
    26

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 45 REFERENCES
    Neural Jump Stochastic Differential Equations
    27
    Variational Inference for Diffusion Processes
    74
    A probabilistic model for the numerical solution of initial value problems
    24
    Neural Ordinary Differential Equations
    638
    Stochastic Backpropagation and Approximate Inference in Deep Generative Models
    2640
    SENSITIVITY ANALYSIS USING ITÔ – MALLIAVIN CALCULUS AND
    51