The recognition of the brain's ability to solve very complex tasks, which are very difficult for conventional methods to solve, has lead to the development of Artificial Neural Network (ANN) technology. ANN technology is very loosely modeled after the massively parallel brain structure and can perform brain like functions such as learning, association, categorization and generalization. Nowadays ANN has been applied in many fields. There are 7 chapters in this thesis: Chapter 1 History of China's Stock Markets and Hong Kong 's Stock Market Chapter 2 Review of the classical method-linear regression Chapter 3 Trading Rules obtained by Regression on Hong Kong, Shenzhen and Shanghai Stock Markets Chapter 4 General concepts of ANN Chapter 5 Feedforward networks Chapter 6 Trading on Stock Markets according to the simulation making by ANN Chapter 7 Comparing the results obtained by Regression and ANN The main purposes of this thesis are: 1. Making trading rules by traditional linear regression 2. Making ANN models applied on stock markets. 3. Show the technique differences between classical Regression and ANN 4. Finding suitable ANN models to make trading on Stock Markets 5. Finding what factors may affect the ANN models 6.Comparsion linear regression method with ANN method