Decision Making Modeling for Stock Portfolio Formation Process through Adaptive Neural Fuzzy Network

This paper focuses on the combination of intelligent, technical and time series techniques. It is necessary to use intelligent series to predict the process of stock price change and deduct the price patterns. In most of the applied intelligent methods, the predictions don't come through. In this project not only the newest intelligent techniques, which are neural and fuzzy nets, are used, but also indicators and time serried regressions have been used simultaneously to increase approximation measure dramatically. This research is aimed at performing an intelligent technique by the help of adaptive neurofuzzy networks and mixing it with time series and technical analysis models. In this way the nonlinear behavior of the stock market, in spite of its rebellious form, can be used as a model. This is the main aim for carrying out this research dramatic error reduction to own the companies stock prices with a suitable approximation.