The present book is focused on a special learning algorithm recently introduced in literature - namely the Repetitive Stochastic Guesstimation. Both analytical methods and numerical simulations are activated, in the effort to place this econometric devoted technique within the frame of Computational Intelligent Techniques and also to underline its apart benefits in the field of short time series econometrics. This approach is justified by the strong demand of a consistent answer to the econometric problems raised by uncertain economic behavior - like the case of countries in transition. It is hard to estimate the impact this kind of research will have on economic theory. Compared to other fields of economic research, developments of this direction have just begun. It is my strong belief that this approach will naturally become a major part of economic research and will offer a new understanding of economic behavior. For this project to be successful two issues are of tremendous importance: having first a theoretical foundation, and second a good linkage to empirical evidence. It was the intention of this book to meet, in equal parts, both these demands.