Short Term Stock Price Forecasting Based On Fuzzy Deep Learning Network Algorithm

The research of Stock market which is a country’s barometer has been an important part of the economics research. The governments and the managers all pay close attention to the short-term stock market compared to the other markets.For a variety of factors, the stock market’s complexity and nonlinearity lead to uncertainty and strong noise of the stock price. So the forecasting stock price result is often unsatisfactory. With the development of the big data, it can provide more information to forecast the stock, at the same time, the data processing power is also an important part in the forecasting market. It is significative and valuable to establish a new theoretical prediction model using the new theory.As a commonly used nonlinear dynamics system, neural network has a wide range of applications in the field of financial prediction problem. The neural network has achieved some good results on academic research. Traditional shallow BP neural network model has some disaduantages, for example, the learning speed is slow and easy to fall into local convergence. The deeping learning neural network which put forward by Hinton professor coming from Canada brings new future in artificial intelligence field. The deeping belief neural which is a typical deeping learning neural model has made a great success in speech recognition and graph extraction field. At the same time, many scholars coming from various fields tried to use the deeping belief neural model in other’s field. The fuzzy theory and neural network application has combined more than forty years. Some theories have proved that the bombination of the two theories can effectively make up for the inadequacy of their own.The market prediction and the theoretical of fuzzy theory and deep belief networks are reviewed in this paper. The article researched the DBN model learning algorithm. This papper constructed a model fuzzy deep learning network share price forecast through the experiment method. it was the first time to use the deep learning network to forecast stock market. In the artical, we analyzed the prediction results and the confirmed the DBN optimal structure through the experiment method. We used the model to forecast different stocks and compared with the BP networks’ results. We analysized the model through different datas incluing the five and ten minutes datas. The price prediction model established use the fuzzy theory and deeping learning network algorithm had a good work which is confirme by the results of the experiments.