Abstract:We present a novel variational framework for performing inference in (neural) stochastic differential equations (SDEs) driven by Markov-approximate fractional Brownian motion (fBM). SDEs offer a versatile tool for modeling real-world continuous-time dynamic systems with inherent noise and randomness. Combining SDEs with the powerful inference capabilities of variational methods, enables the learning of representative function distributions through stochastic gradient descent. However, conventional SDEs typically assume the underlying noise to follow a Brownian motion (BM), which hinders their ability to capture long-term dependencies. In contrast, fractional Brownian motion (fBM) extends BM to encompass non-Markovian dynamics, but existing methods for inferring fBM parameters are either computationally demanding or statistically inefficient. In this paper, building upon the Markov approximation of fBM, we derive the evidence lower bound essential for efficient variational inference of posterior path measures, drawing from the well-established field of stochastic analysis. Additionally, we provide a closed-form expression to determine optimal approximation coefficients. Furthermore, we propose the use of neural networks to learn the drift, diffusion and control terms within our variational posterior, leading to the variational training of neural-SDEs. In this framework, we also optimize the Hurst index, governing the nature of our fractional noise. Beyond validation on synthetic data, we contribute a novel architecture for variational latent video prediction,-an approach that, to the best of our knowledge, enables the first variational neural-SDE application to video perception.

Black-box Variational Inference for Stochastic Differential Equations

Variational Inference for SDEs Driven by Fractional Noise

Variational Inference for Stochastic Differential Equations

Variational Inference for Diffusion Processes.

Variational Inference for Continuous-Time Switching Dynamical Systems

Stochastic Differential Equations with Variational Wishart Diffusions

A New Variational Radial Basis Function Approximation for Inference in Multivariate Diffusions

Dynamic Variational Bayesian Student's T Mixture Regression With Hidden Variables Propagation for Industrial Inferential Sensor Development

Variational Gaussian Process Diffusion Processes

Physics-informed variational inference for uncertainty quantification of stochastic differential equations

Variance Control for Black Box Variational Inference Using The James-Stein Estimator

Black-Box Autoregressive Density Estimation for State-Space Models

Two Methods For Wild Variational Inference

Amortized Reparametrization: Efficient and Scalable Variational Inference for Latent SDEs

Variational Bayesian inference with stochastic search

Parameter Inference for Degenerate Diffusion Processes

Structured Black Box Variational Inference for Latent Time Series Models

Variational Inference for Gaussian Process Modulated Poisson Processes

Variational Markov chain Monte Carlo for Bayesian smoothing of non-linear diffusions

A variational approach to sampling in diffusion processes

Conditional Variational Diffusion Models