Complex Systems Theory Department

20/03/2021

It has already been shown many times that the deterministic model of the financial market evolution based on the discrete scale invariance works well as a tool of forecasting markets' behaviour in more or less distant future. Now we propose a new short-term prediction of the bitcoin price evolution, which should reach its local maximum at the turn of March and April this year. This maximum is expected to end an upward trend on BTC/USD that we have been observing for the last few months. Please note that, in principle, a critical time is the last possible moment of the trend reversal, but it can happen earlier as well. We assume that the critical time is a subject to small adjustments as it can be determined with a finite precision only that is the better the closer is the predicted date. This fact has to be taken into account while following any prediction based on the forementioned model.

07/06/2018

For those, who might be interested in carrying out Ph.D. projects at our department for the following years, we propose a few brand new topics:

1. Application of the categorization based on network local properties to empirical data analysis.
2. Modeling of the mountain range evolution by means of complex network formalism.
3. Exploring linguistic networks through random walk and multifractals.
4. Self-similarity between edges and paths in emergent networks.
5. Synchronization in networks of heterogeneous oscillators.

A brief description of each topic can be found on our web page:

One of the objectives of empirical data analysis is a correct classificastion of the data. In the case of large data sets or in the case of data sets that differ too delicately among themselves, the classification is carried out by means of specific algorithms that often exploit machine learning. Th...

12/02/2016

In a paper published in Physical Review E we proposed an algorithm that is able to properly quantify multifractal structures in time series. We called it multifractal cross-correlation analysis (MFCCA). It is based on the multifractal detrended cross-correlation analysis and it has to replace the algorithms used earlier in literature that were ill-defined and that produced spurious results.
If you have got an access to APS journals: http://dx.doi.org/10.1103/PhysRevE.89.023305 or if you prefer a free preprint: http://arxiv.org/abs/1308.6148 .

We propose an algorithm, multifractal cross-correlation analysis (MFCCA), which constitutes a consistent extension of the detrended cross-correlation analysis and is able to properly identify and quantify subtle characteristics of multifractal cross-correlations between two time series. Our motivati…

12/12/2013

We propose a scenario of the near future evolution of the American stock markets based on the log-periodic model adjusted to the S&P 500 index. According to this model, at the turn of January and February, 2014 we will witness a local reversal/correction of the upward trend that has been observed since 2008. More details can be found at this link: http://www.ifj.edu.pl/dept/no4/nz44/page/modelling-current-sp500.html

19/06/2013

We invite you to visit our new home page at http://www.ifj.edu.pl/dept/no4/nz44/page and find more information about us and our scientific interests, as well as a list of proposed Ph.D. projects that can be done at CSTD. Those visitors who are interested in financial markets are welcome to look at our market prognoses. The corresponding pages will successively be added and extended.