Generating synthetic data for financial time series poses challenges, especially taking into account their non-stationary nature. In this work, we introduce the Sig-Graph Generative Adversarial Network (GAN) model, which integrates the following three components: the time series signature, offering a structured summary of temporal evolution of a times series; a Long Short-Term Memory (LSTM) network, capturing its inherent autoregressive structure; and Graph Neural Networks (GNNs), leveraging geometric patterns within the time series data. Numerical evaluation demonstrates that the Sig-Graph GAN model outperforms several baseline models in replicating the distribution of logarithmic returns over the Standard and Poor’s 500 stock exchanges.
A Generative Adversarial Graph Neural Network for Synthetic Time Series Data
Parton M.
2025-01-01
Abstract
Generating synthetic data for financial time series poses challenges, especially taking into account their non-stationary nature. In this work, we introduce the Sig-Graph Generative Adversarial Network (GAN) model, which integrates the following three components: the time series signature, offering a structured summary of temporal evolution of a times series; a Long Short-Term Memory (LSTM) network, capturing its inherent autoregressive structure; and Graph Neural Networks (GNNs), leveraging geometric patterns within the time series data. Numerical evaluation demonstrates that the Sig-Graph GAN model outperforms several baseline models in replicating the distribution of logarithmic returns over the Standard and Poor’s 500 stock exchanges.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
