The Analysis Of Episodic Memory Model Based On Trade Data Stream In The Stock Market

Stock trade is a kind of investment that takes on high risk forwardly to achieve high proceeds in modern economic movement. Value index of stock only offer a kind of tool for people to weigh the historical change of stock price. Whereas, the prior thing that people cared is how to forecast the coming trend of stock price and the appropriate opportunity to trade. At the same time, the time series of stock price is a nonlinear time series on account of a great many factors which influence the stock price. So, the result of stock forecast is discontented sometimes. On the other hand, Considering a good deal of immature investors dominating our stock exchange and the heterogeneity and changeability of people's reaction on information will be a long-standing phenomenon, So, seeking and researching a kind of scientific and reasonable method to analyze the complexity and rule of the fluctuation of the stock price will be a very important work.In this paper, we design a kind of episodic memory model based on trade data stream in the stock market to seek the historical fact and rule, design an episode by the coming trend of the stock market on one patter, and recall them in the future market, also, they showed the actual historical-phenomena., accordingly investors can choice equally, it is more convictive than any other forecast method, also accord to the rule of the history replay. This model is very convictive to scattered investors, it also can avoid loss resulted by the mistake of informations; The episodic memory model can judge the veracity of techno-index; also can choice the most optimum techno-index for one stock; We also define a method by this episodic memory model to compare the activity of stocks. On the other hand, the techno-index parameter often is an important factor for the efficiency and agility of a techno-index. we also analyze this problem by the result of this episodic memory model, the result indicates that the best techno-index parameter isn't settled, it has something with the stock's trait and investor's love, we can't unionize, we also can't listen the expert's experience blindly, it isn't universal or some one is wrong.