Temporal Visualisation

Abstract

The Generative Topographic Mapping is a probability density model which describes the distribution of data in a space of several dimensions in terms of a smaller number of latent (or hidden) variables. The standard GTM (generative topographic mapping) has been extended to model time series by incorporating it as the emission density in a hidden Markov model. This thesis studies the use of the Generative Topographic Mapping through time model for predicting regime shifts in financial market data. We looked at several aspects of the model, and trained it on different data sets and show the process of quantifying the information in the visualisation plot.

Date of Award	Sept 2002
Original language	English
Awarding Institution	Aston University
Supervisor	Ian Nabney (Supervisor)