Based on the macro factors and open learning´s agent-based model

This paper made some modifications on traditional artificial stock market. It put forward traders´ adaptive learning mechanism, enabling traders to enhance their adaptability to new stock environment by continuous learning. Furthermore, the paper started with external reasons and brought in macrocosmic analysis model, making the artificial stock market closer to the real one. At the same time, validating the characteristics of changes in stock prices, we found they meet EMH basically and the returns took on obvious heavy tail distributions. One stock was selected to be stimulated and forecasted in our artificial stock market, which gained satisfying results.