Qn The Validity Of China's Stock Market From The Perspective Of RBF Neural Network

Basing on the validity theory of the securities market and the extent that the stock prices react to varied data, the market could be divided into three types, which are Strong-form efficient market, semi strong-form efficient market and weak-form efficient market according to the validity in a descending order. The more validity the market is, the more comprehensive and adequate the stock prices reflect the market data. Thus, the stock price is difficult to predict in a market of high validity. Taking the commercial department in the stock industry as the research sample, this paper, sets out to explore the effectiveness of China's stock market by applying the neural network approach.The first chapter of the paper mainly involves two aspects. The first is the research background. It discusses the prediction of the stock price and the accuracy of the prediction methods, the core part of the validity research, which foreshadow the validity research in the following chapter. The second is the review of the research work of validity by researchers from home and abroad, it highlights the researches which applied neural network way to predict stock price and points out their defects, thus to propose the creativity of the paper. The second chapter firstly introduces the definition and categorization of the validity of the stock market, then it highlights several common methods of market validity research. The third chapter introduces the concept of neural network system, from which the characteristics of neural network and its good ability to solve nonlinear problems can be drawn. Then it elaborates the complexity of China's securities market and its nonlinear characteristics, reaching the conclusion that neural network system is a powerful tool to deal with related issues of stock market. Chapter four enters the empirical research part, it first explains the fundamental analysis and related knowledge of RBF networks, then analyzes the data acquisition and processing. The data has to be analyzed by the method of clustering in order to take advantage of RBF network training, drawing the conclusion that whether semi-strong efficient features exist in China's stock market. Chapter five introduces the concepts of technical analysis, and similarly applies RBF network to analyze the acquired data, arriving the conclusion that China's stock market has reached a weak-form efficient state. Chapter six is to make a correlation analysis of the results, attempts to explain features of the degree of the validity by studying the current situation of Chinese stock market. Finally, basing on the research result, in order to further improve the validity of china's stock market, suggestions are proposed, meanwhile, some advice for small and medium investors is provided.