Visually Evaluating Generative Adversarial Networks Using Itself under Multivariate Time Series

TL;DR

This paper introduces a framework called Gaussian GANs for visually evaluating the quality of generated multivariate time series by transforming the data and applying normality tests with chi-square visualization, demonstrating effectiveness on real data.

Contribution

It proposes a novel method to evaluate GAN-generated multivariate time series using Gaussian GANs and normality testing, providing an intuitive visual assessment approach.

Findings

Normality test confirms the quality of generated data

Chi square visualization offers an effective visual evaluation

Empirical results on UniMiB dataset validate the approach

Abstract

Visually evaluating the goodness of generated Multivariate Time Series (MTS) are difficult to implement, especially in the case that the generative model is Generative Adversarial Networks (GANs). We present a general framework named Gaussian GANs to visually evaluate GANs using itself under the MTS generation task. Firstly, we attempt to find the transformation function in the multivariate Kolmogorov Smirnov (MKS) test by explicitly reconstructing the architecture of GANs. Secondly, we conduct the normality test of transformed MST where the Gaussian GANs serves as the transformation function in the MKS test. In order to simplify the normality test, an efficient visualization is proposed using the chi square distribution. In the experiment, we use the UniMiB dataset and provide empirical evidence showing that the normality test using Gaussian GANs and chi sqaure visualization is effective and credible.