Automatic extraction of time-space structure of phase transitions in financial markets
Finished by March 31, 2013
Osaka City University, Graduate School of Business, Associate Professor
Controlling and forecasting phase transition phenomena, such as financial bubbles, is an important issue to be solved. However, given the limited amount of available data which are noisy and with outliers, it is difficult to extract highly complicated structure involving many factors. By using the robust and efficient statistical methods I proposed, this research aims at visualizing the time-space structure of financial bubbles, for the purpose of exploring causal relations and enhancing prediction performance.
