Please use this identifier to cite or link to this item: http://hdl.handle.net/11455/91836
標題: 具動態顆粒與自動標籤之漸進式軌跡資料探勘
Incremental Trajectory Data Mining with Dynamic Granularity and Auto-labeling
作者: Yi-Chen Lu
盧釔辰
關鍵字: Trajectory mining
Auto-labeling
Dynamic granularity
Incremental mining
軌跡資料探勘
自動標籤
動態顆粒
漸進式資料探勘
引用: 參考文獻 [1]Paul Bieganski, John Ned1, John V. Cadis, Ernest E Retzel, “Generalized Suffix Trees for Biological Sequence Data: Applications and Implementation, ” System Sciences, 1994. Proceedings of the Twenty-Seventh Hawaii International Conference on , vol.5, no., pp.35-44, 4-7 Jan. 1994. [2] Zhong, W.; Altum, G.; Harrison, R.; Tai, P.C.; Pan, Y.; , 'Mining protein sequence motifs representing common 3D structures,' Computational Systems Bioinformatics Conference, 2005. Workshops and Poster Abstracts. IEEE , vol., no., pp. 215- 216, 8-11 Aug. 2005. [3] Demiriz, A.; , 'webSPADE: a parallel sequence mining algorithm to analyze web log data,' Data Mining, 2002. ICDM 2003. Proceedings. 2002 IEEE International Conference on , vol., no., pp. 755- 758, 2002. [4]Man Yat Lo; Lucas, S.M.; , 'Evolving Musical Sequences with N-Gram Based Trainable Fitness Functions,' Evolutionary Computation, 2006. CEC 2006. IEEE Congress on , vol., no., pp.601-608, 0-0 0. [5] Greenspan, H.; Goldberger, J.; Mayer, A.; , 'Probabilistic space-time video modeling via piecewise GMM,' Pattern Analysis and Machine Intelligence, IEEE Transactions on , vol.26, no.3, pp.384-396, March 2004. [6] Wang, Chong; Wang, Yanqing; , 'Discovering consumer''s behavior changes based on purchase sequences,' Fuzzy Systems and Knowledge Discovery (FSKD), 2012 9th International Conference on , vol., no., pp.642-645, 29-31 May 2012. [7] Ming-Syan Chen; Jong Soo Park; Yu, P.S.; , 'Efficient data mining for path traversal patterns,' Knowledge and Data Engineering, IEEE Transactions on , vol.10, no.2, pp.209-221, Mar/Apr 1998. [8] Dianhui Wang; Nung Kion Lee; Dillon, T.S.; , 'Data mining for building neural protein sequence classification systems with improved performance,' Neural Networks, 2003. Proceedings of the International Joint Conference on , vol.3, no., pp. 1746- 1751 vol.3, 20-24 July 2003. [9] de Souza Rodrigues, T.; Cardoso, F.C.; Ribeiro Teixeira, S.M.; Oliveira, S.C.; Braga, A.P.; , 'Protein Classification with Extended-Sequence Coding by Sliding Window,' Computational Biology and Bioinformatics, IEEE/ACM Transactions on , vol.8, no.6, pp.1721-1726, Nov.-Dec. 2011. [10]Jingke Xi; , 'Outlier Detection Algorithms in Data Mining,' Intelligent Information Technology Application, 2008. IITA ''08. Second International Symposium on , vol.1, no., pp.94-97, 20-22 Dec. 2008. [11] Koufakou, A.; Secretan, J.; Reeder, J.; Cardona, K.; Georgiopoulos, M.; , 'Fast parallel outlier detection for categorical datasets using MapReduce,' Neural Networks, 2008. IJCNN 2008. (IEEE World Congress on Computational Intelligence). IEEE International Joint Conference on , vol., no., pp.3298-3304, 1-8 June 2008. [12] Zhipeng Cai; Lizhe Xu; Yi Shi; Salavatipour, M.R.; Goebel, R.; Guohui Lin; , 'Using Gene Clustering to Identify Discriminatory Genes with Higher Classification Accuracy,' BioInformatics and BioEngineering, 2006. BIBE 2006. Sixth IEEE Symposium on , vol., no., pp.235-242, 16-18 Oct. 2006. [13] Nagar, A.; Al-Mubaid, H.; , 'Using path length measure for gene clustering based on similarity of annotation terms,' Computers and Communications, 2008. ISCC 2008. IEEE Symposium on , vol., no., pp.637-642, 6-9 July 2008. [14]Chen, X.; Murphy, R.F.; , 'Location proteomics: determining the optimal grouping of proteins according to their subcellular location patterns as determined from fluorescence microscope images,' Signals, Systems and Computers, 2004. Conference Record of the Thirty-Eighth Asilomar Conference on , vol.1, no., pp. 50- 54 Vol.1, 7-10 Nov. 2004. [15] Chih-Chieh Hung; Wen-Chih Peng; , 'A Storage Management for Mining Object Moving Patterns in Object Tracking Sensor Networks,' Web Intelligence and Intelligent Agent Technology Workshops, 2006. WI-IAT 2006 Workshops. 2006 IEEE/WIC/ACM International Conference on , vol., no., pp.252-258, Dec. 2006. [16] Gaol, F.L.; , 'Exploring the Pattern of Habits of Users Using Web log Squential Pattern,' Advances in Computing, Control and Telecommunication Technologies (ACT), 2010 Second International Conference on , vol., no., pp.161-163, 2-3 Dec. 2010. [17] Chih-Chin Liu; Jia-Lien Hsu; Chen, A.L.P.; , 'Efficient theme and non-trivial repeating pattern discovering in music databases,' Data Engineering, 1999. Proceedings., 15th International Conference on , vol., no., pp.14-21, 23-26 Mar 1999.
摘要: 軌跡型樣探勘(Trajectory Pattern Mining) 主要目的就是從軌跡序列資料中挖掘出隱藏其中的特殊、重要且具代表性的移動行為和特徵(feature),是資料探勘非常熱門的領域之一。探勘所得的移動型樣(Movement Pattern)應用性很廣,可以用於移動軌跡的預測或相似度的比對,或進一步發展許多有趣的應用。 許多軌跡資料探勘的方法會將監控區域切割成網格,藉以將物件移動軌跡的經緯度座標點序列,轉成一連串的網格序列,以利軌跡型樣的探勘。但這樣的方法可能面臨的問題包括下列五方面: 一、不容易找出一個最佳的網格大小與切割位置。二、網格數量可能很大,致使探勘耗費大量的CPU或記憶體。三、採用client-server架構,將軌跡資料上傳到server端做探勘再回傳結果的策略有隱私洩漏的疑慮,不利相關應用的發展。四、大量的軌跡資料導致傳統演算法的計算時間增加,甚至無法執行。五、傳統的型樣定義和演算法無法快速反應移動行為的變化。基於上述五大問題和挑戰,本論文提出Incremental PST with Dynamic Granularity and Auto-labeling (incPSTDGAL )演算法,其根據個人的移動習慣,動態決定每個區域的網格大小,找出具代表性的網格,並分配重要網格label,以減少總label的總數,再採取incremental mining的策略,來因應不斷累積增加的軌跡資料;也就是將累積的資料分批探勘,再將每次探勘的結果合併,如此可以避免一次探勘很大量的資料,此外,我們設計Aging的調整機制,在合併探勘結果時,給予新的探勘結果較大的weighting,以利快速反應使用者行為的劇烈變動。這一系列的改進使單次軌跡資料探勘的計算時間大幅下降,有助於直接在客戶端直接進行個人軌跡資料的探勘,無形中也加強了使用者隱私保護。 為驗證incPSTDGAL演算法的效能,我們於Android手機平台實作演算法,並設計多個實驗來比較演算法的效率和探勘結果的有效性。經實驗證實,incPSTDGAL演算法克服傳統PST的限制,可以有效控制每次探勘的計算成本,因此,我們可以在client端做軌跡資料探勘而不需將資料上傳到server端,且因為動態調整有效網格的顆粒粗細,使探勘所得的移動型樣更多更有效。
URI: http://hdl.handle.net/11455/91836
文章公開時間: 10000-01-01
Appears in Collections:通訊工程研究所

文件中的檔案:

取得全文請前往華藝線上圖書館



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.