Component analysis in financial time series
R.H. Lesch, Y. Caille, David G. Lowe · 2003
We discuss the application of principal component analysis and independent component analysis for blind source separation of univariate financial time series. In order to perform single-channel versions of these techniques, we work within the embedding framework, using delay coordinate vectors to obtain a multidimensional representation of the system dynamics at each time instance. The main objective is to find out if these techniques are able to perform feature extraction, signal-noise-decomposition and dimensionality reduction, since that would enable a further inside look into the behaviour and mechanics of financial markets. Both methods are applied to the currency exchange rate data of the British Pound against the US Dollar.