Abstract
This thesis explores temporal structure based on selfsimilarity in different contexts.
An efficient dynamic programming algorithm is presented which discovers temporal structures in music shows, obtains high quality results, and compares them to similar algorithms used in the literature. The program segments a selfsimilarity matrix given a cost function and a fixed number of homogeneous
temporal structures to find. This is the initial approach we use to discover temporal structures in music data.
The use of a selfsimilarity matrix to visualize temporal structures is discussed in detail. Then the following question is explored; if similar temporal structures in other corpora existed; could forecasting algorithms be adapted to take advantage of them even if they were not known a priori?
Prediction with expert advice techniques are then introduced to exploit a priori unknown temporal structures of a similar configuration in an online configuration. Univariate Russian Stock Exchange options futures volatility corpora are used, which are highly interesting for online forecasting.
We experiment with merging together expert models which have been trained in some way to recognise temporal structures in corpora. The first types are kernel ridge regression models trained to be experts on particular regions in time, or
untrained and given random sets of parameters which may work well on certain time regions. The other types of model used are parsimonious predictors which transform univariate financial data into elementary time series based on homogeneous vicinities of information in the side domain. Expert merging techniques are then used across these time series which produce a validationfree forecaster comparable to sliding kernel ridge regression.
An efficient dynamic programming algorithm is presented which discovers temporal structures in music shows, obtains high quality results, and compares them to similar algorithms used in the literature. The program segments a selfsimilarity matrix given a cost function and a fixed number of homogeneous
temporal structures to find. This is the initial approach we use to discover temporal structures in music data.
The use of a selfsimilarity matrix to visualize temporal structures is discussed in detail. Then the following question is explored; if similar temporal structures in other corpora existed; could forecasting algorithms be adapted to take advantage of them even if they were not known a priori?
Prediction with expert advice techniques are then introduced to exploit a priori unknown temporal structures of a similar configuration in an online configuration. Univariate Russian Stock Exchange options futures volatility corpora are used, which are highly interesting for online forecasting.
We experiment with merging together expert models which have been trained in some way to recognise temporal structures in corpora. The first types are kernel ridge regression models trained to be experts on particular regions in time, or
untrained and given random sets of parameters which may work well on certain time regions. The other types of model used are parsimonious predictors which transform univariate financial data into elementary time series based on homogeneous vicinities of information in the side domain. Expert merging techniques are then used across these time series which produce a validationfree forecaster comparable to sliding kernel ridge regression.
Original language  English 

Qualification  Ph.D. 
Awarding Institution 

Supervisors/Advisors 

Award date  1 Apr 2015 
Publication status  Unpublished  2015 
Keywords
 Machine Learning
 Online regression
 time series
 music
 dj
 finance
 forecasting
 temporal structures
 regions
 regimes
 prediction with expert advice
 segmentation
 merging
 experts