Research Of Time Series Data Mining And Visualization

The method of Time Series Data Mining is used in meteorology, finance, medicine, electricity, hydrology, industrial control and many other fields widely. It has important value on research. Traditional time series data mining technologies use the method of statistics and artificial annual networks. In recent years, regression analysis becomes the most popular method of mathematical statistics method gradually. This method can solve the problem of relationship among variables. And solve prediction, control, optimization and other problems. There are two types in regression algorithm, linear regression algorithm and nonlinear regression algorithm. Nonlinear regression algorithm must construct a regression model made by the symbolic expression first, the process of regression is the process of determine the expression coefficient.The main work and conclusions of this paper are presented as follows:1. We introduce the birth background and main method of the data mining technology. Then we introduce the techniques and methods of visualization. The final outcome of prediction is given to the user, it is particularly important to show to users using a visual method of the results. In this paper, we not only introduce the visualization method, but also on the important significance and application prospects of development of visualization.2. We introduce the concept of time series and time series analysis. Then we describe the current research level of time series data mining.3. We elaborate support vector regression method and the support vector machine that can be used for function fitting.4. We introduce the classification of support vector regression algorithm, and the algorithm is designed on the basis of a time series prediction model and proposed a simple but effective model for computing parameters. Then we give the feasibility experiment of the prediction model. The result of the experiment verifies that the method is available and valuable. At the end of this paper, the result of the experiment is visualized, and verifies the feasibility of time series visualization based on CSVR.