Research of new learning method of agent to predicting stock performance

Predicting stock performance is a very large and profitable area of study. A large number of studies have been reported in literature with reference to the use of artificial neural network in modeling stock performance in western countries. However, not much work along the approach to neural network based on agent has been reported. This thesis focuses on the development and the simulation of a stock market performance model of utilizing a neural network approach base on agent. This model is easy to understand, and can be easily implemented as a software simulation. First we will discuss the basic concepts behind this type of neural network based on agent, then, we´ll get into some of the more application ideas.