Stock Markets Prediction

Stock markets data became very important object of academic research. Failures of big market players and financial supervision during the financial crisis showed that standard time series models have several shortages in precision and robustness. Most of the conventional techniques have been trying to capture the patterns in the examined data using linear relationships and assumptions. But because there is no empirical evidence of linearity in stock returns, various researchers and financial practitioners have focused on the nonlinear prediction methods. Artificial neural networks are computational structures that are capable of accurate forecasts without any specific assumptions about the distribution or characteristics of the observed variables and their effectiveness has been proved in many real world applications. The main objective of this book is to introduce a new G-S network that combines the essential principles of standard neural networks with metaheuristics, represented by genetic algorithms and simulated annealing. It is shown that proposed hybrid network clearly outperforms standard techniques as well as ordinary neural networks