A computational approach in dealing with uncertainty of financial markets

Financial markets play a very important role in our life. During last decades a substantial progress has been made in their understanding. A main result of these developments is the change from random walk models to models that view the financial market as a complex dynamical system. Recently, it is discovered that financial markets have non-random modes.

Non-random modes are significant in dealing with uncertainty of financial markets. However, understanding of non-random modes is limited and there is no fundamental theory about them in general. Currently, intensive investigations are underway to change the situation. A contribution to the solution of these problems is made in the thesis. In particular, a computational approach to characterise and quantify non-random modes of the financial market is developed. The development is realised to a stage where the approach may be tested and used for real data of financial markets.