Year: 2024
Type: dissertation
Abstract: This thesis investigates the use of neural network models for time series forecasting \nwith an emphasis on modelling latent dynamics (unobservable time series). \nTime series forecasting is of great research interest to both industry and \nacademia. It describes the task of predicting the future values of one or more time \nseries conditioned on past information. With growth in the availability of data, researchers have started developing machine learning - more specifically neural networks - to model more complex temporal dynamics. Neural networks excel at modelling non-linearity and requires fewer assumptions on the underlying process. \nTo contribute to the area of deep learning for time series forecasting, I focus \nspecifically on the fact that not all time series can be observed. Financial markets \nfor example, contain unobservable market regimes which drive observed time series \nsuch as stock returns and prices. Being able to model these latent dynamics provide \nus with a useful tool to study what is happening beneath the surface. To achieve this, \nI bring topics together from other areas of machine learning such as representation \nlearning and Bayesian inference. \nThis thesis is broken down into four experiments. In the first experiment, I \ndevelop a stochastic variant of the recurrent neural network which can be used to \nperform multi-step-ahead time series forecasting, and generate confidence intervals \nfor the predictions. \nIn the second experiment, I study the concept of Granger causality in the presence of a potential confounder. I develop a neural network architecture to model \nthe confounder and show that by taking this into account, one can obtain better \nforecasting accuracy on the target time series. The third and fourth experiments are concerned with the application of latent variable modelling in financial markets. In experiment three, I bring together deep learning and GARCH models from the field of econometrics and propose a neural architecture for volatility (variance/covariance) forecasting in a low dimensional setting (<5 assets). Finally, in the fourth experiment, I build on my work in the third experiment to propose a model capable of forecasting the volatility of an investment portfolio in higher dimensional settings.
Author: Yin, Zexuan
Institution: University College London
Language: English
Field: Decision Sciences
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7021196614