The Research Of How To Prediction The Internet Company’s Stock Price

In recent years, with the rapid development of economics and the perfecting of stock market, stock has become an important part of people’s investment. After experiencing the network economy foam, internet companies are more and more mature, and meanwhile with the character of efficient and low cost, they also make a great impact on traditional enterprises. Internet companies are high-growth and high-risk, the analysis to their valuation and stock has get much more attention from venture capitalists, especially general investors. So an effective evaluate to internet company valuation and stock price has theoretical significance and applicable value for investors.In terms of valuation about internet company, this paper has discussed the development status and prospects mainly from political、economics、national policy and other perspectives. What’s more, company profitability、company visibility、 market concentration and innovation capacity, market share and generalization ability, management team and financing capacity and other indexes are used to evaluate the valuation of internet company. Among the specific methods of company valuation, there are some important methods such as Price Earning Radio (PER) model, Economic Value Added model, Discounted Cash Flow model and other assessment model. As this paper doesn’t focus on this study, so they will not be elaborated.In the prediction method of company stock, after comparing various traditional stock forecasting methods, neural network forecasting method is presented. Based on studying these existing problems of standard BP algorithms in stock forecasting, including the slow learning speed, local extremum and the low prediction precision, this method improves the efficiency and the accuracy of the results by adjusting the activation function of neurons and implementing normalization process of the sample data.As this paper discussing internet company, so make a introduction about definition、classify and character of internet company first, then introducing the basic knowledge about stock and artificial neural network. What’s more, analysis the political and economic influencing factors on internet industry, and through the analyzing about internet company valuation’s source and the main factor of affecting the valuation, in order to have a pre-judgment on stock investment and make a reasonable evaluation to internet stock.By analyzing traditional stock forecasting methods, such as securities investment approach and model prediction method, indicating the unique character of neural network prediction and forecasting the feasibility of stock trend. The neural network prediction model has established by designing topology of BP Neural network and the selecting、classifying、normalization processing on sample data, then with collected data doing simulation prediction experiment by Matlab simulation software.Based on the raw data during May2011and August2012stock from internet company (Focus Technology) listed in domestic market, calling BP Neural network model which has adjusted to learn and train, by comparing the former and latter results through activity function, the accuracy and efficiency of the latter forecasting model improve dramatically.The analysis from macro and micro enable investors catch economics hot pots effectively when they are investing the internet company and provide a theoretical basis and data support for investors in order to get maximize return on investment and minimize risk. The stock market is a very complex dynamic system, so the predicted results with the target results may be out under the impact of personal factors、external political and economic factors and so on. In addition, due to limited in personal capacity, there is a certain lack in the improvement of the neural network algorithm, which need further explore and study. Combining this powerful simulation software strive to perfect a neural network toolbox suitable for internet company stock forecasting.