Font Size: a A A

Design And Implementation Of The Algorithms On Time Series Analysis And Forecast

Posted on:2009-11-03Degree:MasterType:Thesis
Country:ChinaCandidate:K YanFull Text:PDF
GTID:2189360245969315Subject:Software engineering
Abstract/Summary:PDF Full Text Request
This article discusses the main problem of time series analysis in the field of forecast application and programming.This paper presents the basic theory of the time series analysis, at the same time introduced the use of time-series modeling and forecasting methods in predicting the Ministry of Public Security and its database of crime trends in the application. The so-called time series prediction is forecast to reflect the changes of things past and present structure of the data time series model. With the model and then extrapolated to predict the future. Adopting the parameter identification method and combined with the dynamic data of sedimentation of crime, the time series prediction model of deformation of crime data has been established to monitor and forecast the deformation. The examples of application has shown that this model can monitor and forecast the trends of crime with accuracy, providing important practical application value.The results of this research are to establish social safety incidents in the Ministry of Public Security forecast and early warning system. The papers put forward not only in time series analysis and forecasting of mathematics solutions, while the completion of the algorithm in the programming on the design and implementation. The project using object-oriented programming ideas, from the actual demand of users, in accordance with the general development of software engineering processes, time series analysis to predict the statistical algorithms as the basis, in accordance with the needs analysis, the outline design, detailed design, coding , model testing the general steps to complete the entire project development and completion of the project code embedded in the system through to the actual practical application to verify its feasibility and correctness. The paper focused on the time-series model of model identification method and the establishment of model equations method, but also focuses on these methods in object-oriented programming in the process of achieving the specific way. Finally, the article on the work of the project concluded.This paper includes the following aspects of technology:Probability aspect, the main use is time series analysis methods, including smooth series of judgment, smooth treatment, time series model identification, the judgment of model and the model parameters of the sequence estimation methods. In addition, there are indicators of the seasonal time series analysis, error handling techniques of model analysis; the aspect of Linear algebra, the main use is the knowledge of matrix, including the matrix to the home, the inverse matrix, the matrix of positive features, such as matrix multiplication. In addition, it involves a high-linear equations to solve; about Object-oriented programming techniques, mainly related to the object-oriented programming model of the design, the various functional types of design and preparation and C # programming language in the basic grammar and basic structure of the array of processing, graphics processing, and other relevant Technical knowledge.
Keywords/Search Tags:Time Series, Object-Oriented Programming, Anna Model, Prediction, Parameter Estimates, Model Recognition
PDF Full Text Request
Related items