Nansde-net: A neural sde framework for generating time series with memory

TL;DR

NANSDE-Net introduces a neural network-based noise process for neural SDEs, enabling effective modeling of time series with long- and short-memory effects within the Itô calculus framework, outperforming existing methods.

Contribution

The paper proposes NA-noise, a novel neural network-parameterized Itô-process capable of capturing memory effects, and develops NANSDE-Net, a generative model extending Neural SDEs with this noise.

Findings

NANSDE-Net effectively models long- and short-memory time series.

It outperforms fractional SDE-Net in empirical tests.

Theoretical guarantees of solution existence and uniqueness.

Abstract

Modeling time series with long- or short-memory characteristics is a fundamental challenge in many scientific and engineering domains. While fractional Brownian motion has been widely used as a noise source to capture such memory effects, its incompatibility with It\^o calculus limits its applicability in neural stochastic differential equation~(SDE) frameworks. In this paper, we propose a novel class of noise, termed Neural Network-kernel ARMA-type noise~(NA-noise), which is an It\^o-process-based alternative capable of capturing both long- and short-memory behaviors. The kernel function defining the noise structure is parameterized via neural networks and decomposed into a product form to preserve the Markov property. Based on this noise process, we develop NANSDE-Net, a generative model that extends Neural SDEs by incorporating NA-noise. We prove the theoretical existence and uniqueness of the solution under mild conditions and derive an efficient backpropagation scheme for training. Empirical results on both synthetic and real-world datasets demonstrate that NANSDE-Net matches or outperforms existing models, including fractional SDE-Net, in reproducing long- and short-memory features of the data, while maintaining computational tractability within the It\^o calculus framework.