The book is devoted to the analysis of big data in order to extract from these data hidden patterns necessary for making decisions about the rational behavior of complex systems with the different nature that generate this data. To solve these problems, a group of new methods and tools is used, based on the self-organization of computational processes, the use of crisp and fuzzy cluster analysis methods, hybrid neural-fuzzy networks, and others. The book solves various practical problems. In particular, for the tasks of 3D image recognition and automatic speech recognition large-scale neural networks with applications for Deep Learning systems were used. Application of hybrid neuro-fuzzy networks for analyzing stock markets was presented. The analysis of big historical, economic and physical data revealed the hidden Fibonacci pattern about the course of systemic world conflicts and their connection with the Kondratieff big economic cycles and the Schwabe-Wolf solar activity cycles. The book is useful for system analysts and practitioners working with complex systems in various spheres of human activity.

This monograph is dedicated to the systematic presentation of main trends, technologies and methods of computational intelligence (CI). The book pays big attention to novel important CI technology- fuzzy logic (FL) systems and fuzzy neural networks (FNN). Different FNN including new class of FNN- cascade neo-fuzzy neural networks are considered and their training algorithms are described and analyzed. The applications of FNN to the forecast in macroeconomics and at stock markets are examined. The book presents the problem of portfolio optimization under uncertainty, the novel theory of fuzzy portfolio optimization free of drawbacks of classical model of Markovitz as well as an application for portfolios optimization at Ukrainian, Russian and American stock exchanges. The book also presents the problem of corporations bankruptcy risk forecasting under incomplete and fuzzy information, as well as new methods based on fuzzy sets theory and fuzzy neural networks and results of their application for bankruptcy risk forecasting are presented and compared with Altman method. This monograph also focuses on an inductive modeling method of self-organization - the so-called Group Method of Data Handling (GMDH) which enables to construct the structure of forecasting models almost automatically. The results of experimental investigations of GMDH for forecasting at stock exchanges are presented. The final chapters are devoted to theory and applications of evolutionary modeling (EM) and genetic algorithms. The distinguishing feature of this monograph is a great number of practical examples of CI technologies and methods application for solution of real problems in technology, economy and financial sphere, in particular forecasting, classification, pattern recognition, portfolio optimization, bankruptcy risk prediction under uncertainty which were developed by authors and published in this book for the first time. All CI methods and algorithms are presented from the general system approach and analysis of their properties, advantages and drawbacks that enables practitioners to choose the most adequate method for their own problems solution.

This monograph is dedicated to the systematic presentation of main trends, technologies and methods of computational intelligence (CI). The book pays big attention to novel important CI technology- fuzzy logic (FL) systems and fuzzy neural networks (FNN). Different FNN including new class of FNN- cascade neo-fuzzy neural networks are considered and their training algorithms are described and analyzed. The applications of FNN to the forecast in macroeconomics and at stock markets are examined. The book presents the problem of portfolio optimization under uncertainty, the novel theory of fuzzy portfolio optimization free of drawbacks of classical model of Markovitz as well as an application for portfolios optimization at Ukrainian, Russian and American stock exchanges. The book also presents the problem of corporations bankruptcy risk forecasting under incomplete and fuzzy information, as well as new methods based on fuzzy sets theory and fuzzy neural networks and results of their application for bankruptcy risk forecasting are presented and compared with Altman method. This monograph also focuses on an inductive modeling method of self-organization - the so-called Group Method of Data Handling (GMDH) which enables to construct the structure of forecasting models almost automatically. The results of experimental investigations of GMDH for forecasting at stock exchanges are presented. The final chapters are devoted to theory and applications of evolutionary modeling (EM) and genetic algorithms. The distinguishing feature of this monograph is a great number of practical examples of CI technologies and methods application for solution of real problems in technology, economy and financial sphere, in particular forecasting, classification, pattern recognition, portfolio optimization, bankruptcy risk prediction under uncertainty which were developed by authors and published in this book for the first time. All CI methods and algorithms are presented from the general system approach and analysis of their properties, advantages and drawbacks that enables practitioners to choose the most adequate method for their own problems solution.