Year: 2024
Type: dissertation
Abstract: Good quality data are key to informed decision making. Yet real-world data collection \ncan be challenging and resource demanding, so that good quality data are not always \navailable. This thesis explores synthetic data generation, focusing on Generative \nAdversarial Networks (GANs) which are a class of machine learning models. Real-life \nenvironmental data such as water quality data brings added challenges in that they are \noften datasets with missing values, time dependencies and rare events. The aim of the \nthesis was to investigate the potential of GANs to synthesise such data. \nIdentification of the data synthesis techniques available for such datasets showed that \nwhile GANs are complex models with multiple parameters to optimise, they could be well \nsuited for this purpose. To address the challenge of synthetising unbalanced irregular time \nseries, first a novel GAN was built to create synthetic data directly from datasets with \nmissing values. Called MaWGAN, it is based on the Wasserstein distance (like WGAN) \nand uses a mask to hide missing values from the Critic to enable the GAN to run. \nThen to capture the dependency in the data, two routes were developed: a \nnew algorithm (Force-GAN) that embedded a third neural network in the MaWGAN \narchitecture, and an alternative option where missing data in the irregular time series are \nfirst imputed with a novel technique (Hankel Imputation) that preserves the noise in the \ndata. These insights are then used to illustrate the value of the novel GANs developed \nin this thesis to address complex real-world challenges such as taste and odour issues in \ndrinking water. \nResearch contributions to the field include: a GAN that can handle missing values in the dataset (MaWGAN), an expansion that handles time series as well (Force-GAN), a method to impute data while keeping the original noise and trend of the dataset (Hankel Imputation). The value of these novel methods to predict rare events in time series is also demonstrated using water quality unbalanced datasets with missing values.
Author: Poudevigne-Durance, Thomas
Institution: Cardiff University
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7056494773
