Time Series Data Mining

With the popularity of computer and information technology,and the great development of storage technique of high capacity,,a great amount of data is accumulated in daily work and in scientific research.Much potentially useful knowledge is hided behind data.Today how to manage and use these time series data efficiently and extract useful information is an important problem in data mining.Time is the inherent attribute of data,so we should take time into account when mining association rules.Time series data are a kind of important data existing in a lot of fields, such as financial market,industrial process,science experi-me nts,etc,and the quantity of time series data has explosively increas.So it is necessary to study on the subject of time series data mining.Beginning at the end of 20m century,the research of complex network has been permeating through different areas,such as life science,mathematical science,engineering science,social science and so on.It becomes a very important and challenged subject in scien- tific research.Detecting community structure becomes one of the most popular issues in complex networks.In fact, searching for communities in network is equivalent to clustering data. The new algorithm of complex networks for time series clustering is important.The following work has been done in this dissertation by combination of time series data mining and complex network theory:Summarizes status of the time series data mining and sequence pattern mining in time series firstly,and then summarized the general methods of sequence pattern mining. Summarized the general methods of sequence pattern mining.Propose an algorithm based on Chun-Hao Chen et al.approach tranform the fuzzy frequent trends mining into sequence patt- ern mining,and use the GSP to generated the candidate pattern sets and the pruning is also performed to remove redundant patterns for the mining of the frequent trends efficien- tly .Finally,experiments are also made for different parameter settings,with experimental results showing that the proposed algorithm can actually work.A new algorithm is proposed to fast detect community structures in complex network. The node with the highest degree and its neighbors in complex network are firstly found, and then its neighbor matrix and density set are constructed. All community structures will be obtained by repeating this process. Using of the local information will lead to the reduction of running time. The experiment results show the validity of the algorithm. Furthermore, the proposal is also applied to solve other problems in data mining, such as time series clustering ,etc.