Conditional GAN for timeseries generation

TL;DR

This paper introduces TSGAN, a new GAN architecture designed for realistic one-dimensional time series generation, especially effective with limited data, outperforming existing methods on multiple datasets.

Contribution

The paper proposes TSGAN, a novel architecture tailored for time series data, demonstrating improved performance with limited data compared to existing models.

Findings

TSGAN outperforms competing models on benchmark datasets.

TSGAN achieves higher FID scores indicating better quality.

Qualitative evaluation shows improved realism in generated data.

Abstract

It is abundantly clear that time dependent data is a vital source of information in the world. The challenge has been for applications in machine learning to gain access to a considerable amount of quality data needed for algorithm development and analysis. Modeling synthetic data using a Generative Adversarial Network (GAN) has been at the heart of providing a viable solution. Our work focuses on one dimensional times series and explores the few shot approach, which is the ability of an algorithm to perform well with limited data. This work attempts to ease the frustration by proposing a new architecture, Time Series GAN (TSGAN), to model realistic time series data. We evaluate TSGAN on 70 data sets from a benchmark time series database. Our results demonstrate that TSGAN performs better than the competition both quantitatively using the Frechet Inception Score (FID) metric, and qualitatively when classification is used as the evaluation criteria.