The purpose of this study is to develop two approaches in economic modelling based on the features of the available economic data. The first approach considers economic environments with data of low quality and short time series. This is usual for countries in transition. The presented model not only uses mainstream macroeconomic theories, but also focuses on the particularities of the economy of Azerbaijan, which is the case study in this work. The simulations of the model are consistent with the recent facts of economic development in Azerbaijan. The second part of the book, considers economic environment with qualitative, consistent and highly descriptive data. We propose to take advantage of this fact and develop methods, which are based on artificial intelligence concepts. These methods are largely free of theoretical assumptions and objectively use internal knowledge of data. They incluide a system based on the association rules, a genetic algorithm-fuzzy logic system, a fuzzy-neural architecture and, finally, a hybrid system. Implementations of these methods produce better results than traditional statistical methods and neural networks do alone.