学校代号 10532 学 号 LB2010035 分 类 号 TP391 密 级 普通 博士学位论文 时间序列分析技术的研究时间序列分析技术的研究 (英文版英文版) 学位申请人姓名 VO THI THANH VAN 培 养 单 位 信息科学与工程学院 导师姓名及职称 骆嘉伟 教授 学 科 专 业 计算机应用技术 研 究 方 向 数据挖掘和知识发现 论文提交日期 2013-06-07 University ID : 10532 Student ID : LB2010035 Security Level : Normal Hunan University PhD Thesis THE RESEARCH ON TIME SERIES ANALYSIS TECHNIQUES Applicants Name : VO THI THANH VAN College : Information Science and Engineering Supervisor : Professor Luo Jiawei Major : Computer Science Research Field : Data Mining and Knowledge Discovery Submission Date : 2013-05 Defense Date : 2013-06-07 Defense Committee Chairman : Professor Bo Liao The Research on Time Series Analysis Techniques By VO THI THANH VAN Masters in Computer Science (University of Science, Viet Nam National University of Ho Chi Minh City) 2004 A dissertation submitted in partial satisfaction of the Requirements for the degree of Doctor of Science in Computer Science in the Graduate School of Hunan University Supervisor Professor Luo Jiawei June, 2013I HUNAN UNIVERSITY DECLARATION I, VO THI THANH VAN, hereby declare that the work presented in this PhD thesis titled “The Research on Time Series Analysis Techniques”, is my original work and has not been presented elsewhere for any academic qualification. Where references have been used from books, published papers, reports and web sites, it is fully acknowledged in accordance with the standard referencing practices of the discipline. Students signature:. Date: . Copyright Statement Permission is herewith granted to Hunan University to circulate and reproduce for non-commercial purposes, at its discretion, this thesis upon the request of individuals or institutions. The author does not reserve other publication rights and the thesis nor extensive extracts from it be printed or otherwise reproduce without the authors written permission This thesis belongs to: 1. Secure , and this power of attorney is valid after 2. Not secure . (Please mark the above corresponding check box with“”) Authors Signature: . Date: . Supervisors Signature: Date: . The Research on Time Series Analysis Techniques II ABSTRACT Data mining is the analysis of observed data sets in order to find the models and to summarize the data in the new ways that are meant for both understandable and useful. Data arriving in time order arises in fields ranging from many other areas of physics, finance, medicine, music, and so on. The time series is an important class of temporal data objects and they can be easily obtained from financial and scientific applications. Time series analysis comprises methods and techniques for analyzing time series data in order to extract meaningful statistics and other characteristics of the data. Given the spread of the appearance of time series data, and the exponentially growing sizes of databases, there has been recently been an explosion of interest in time series data mining. As extremely large time series data sets grow more prevalent in a wide variety of settings, this thesis faces the significant challenge of developing efficient analysis methods. The researches in this thesis address the problem in designing fast, scalable algorithms for the analysis of time series. The research on time series analysis with the tasks such as preprocessing and transformation data for the prediction purpose has a meaningful and popular in the case of big size data. If the data or time series data in particular can be preprocessed so as to improve the efficiency and lack of difficulty of the mining and discovering processes. There are a lot of data preprocessing data techniques; to remove the noise and correct incompatibilities in data, the cleaning techniques can be applied; to merge data from multi sources into coherent data storage, the integration techniques can be used; to normalize data, the transformation techniques can be referred. Data reduction is one of the meaningful techniques in the preprocessing stage of time series analysis can reduce the data size by collecting, eliminating redundant features. In general, time series predictability is a measure of how well future values of a time series can be predicted, where a time series is a sequence of observations. Time series predictability indicates to what extent the past can be used to determine the future in a time series. A time series generated by a deterministic linear process has high predictability, and its future values can be predicted very well from the past values. A time series generated by an uncorrelated process has low predictability, and its past values provide only a statistical characterization of the future values. This thesis makes four major contributions: