TSAM - Time Series Analysis and Mining for Financial Risk Management

The project aims to apply, adapt and compare a wide-range of techniques from the fields of artificial intelligence, econometrics, statistics and information theory to financial risk management, and in particular to portfolio selection.

Advanced statistical models that accurately model financial returns and stochastic volatilities will be studied and compared to traditional econometric time series models. Furthermore, different methods and algorithms for the analysis and prediction of financial time series will be studied and compared.

Our aim will be to discover inherent structures and qualitative relationships in financial multivariate time series and generate an intelligible and characterizing description of the discovered patterns. For this modern AI-based and statistical methods will be used and compared. An improvement of the quality of prediction is expected.

Funding entity: PRAXIS.