Danilchuk H. Entropy analysis and modelling of complex socio-economic systems

The thesis offers systematic spectrum entropy indicators, based on which a number of measures of complexity are built, enabling classification of complex systems by type of dynamics. Entropy indicators calculating method has been elaborated. This method allows to investigate original time series, as well as its transformed variants (returns, volatility, fluctuations, etc.). The received multiscale entropy indicators models enable to evaluate the complexity of the system both in space and time and change the rules of socio-economic systems regulation. The process of socio-economic systems monitoring and crisis forecast by using entropy indicators and elaboration of leading indicators of a crisis have been improved, which enables to receive information about negative phenomena in systems in advance and thus to prevent losses. The method for calculating the wavelet entropy for a moving window, which offers an interpretation of specific combinations of waves has been taken further, which in its turn makes it possible to get advance information about possible complex system's dynamics and trends changes and develop crisis response program in advance. The method of estimation of socio-economic systems non-extensive complexity measure which takes into account nonadditive component of Tsallis entropy has also been taken further.